Latent Space: The AI Engineer Podcast - The Future of Email: Superhuman CTO on Your Inbox As the Real AI Agent (Not ChatGPT) — Loïc Houssier

Episode Date: December 11, 2025

From applied cryptography and offensive security in France’s defense industry to optimizing nuclear submarine workflows, then selling his e-signature startup to Docusign (https://www.docusign.com/company/news-center/opentrust-joins-docusign-global-trust-network and now running AI as CTO of Superhuman Mail (Superhuman, recently acquired by Grammarly https://techcrunch.com/2025/07/01/grammarly-acquires-ai-email-client-superhuman/), Loïc Houssier has lived the full arc from deep infra and compliance hell to obsessing over 100ms product experiences and AI-native email. We sat down with Loïc to dig into how you actually put AI into an inbox without adding latency, why Superhuman leans so hard into agentic search and “Ask AI” over your entire email history, how they design tools vs. agents and fight agent laziness, what box-priced inference and local-first caching mean for cost and reliability, and his bet that your inbox will power your future AI EA while AI massively widens the gap between engineers with real fundamentals and those faking it.We discuss:* Loïc’s path from applied cryptography and offensive security in France’s defense industry to submarines, e-signatures, Docusign, and now Superhuman Mail* What 3,000+ engineers actually do at a “simple” product like Docusign: regional compliance, on-prem appliances, and why global scale explodes complexity* How Superhuman thinks about AI in email: auto-labels, smart summaries, follow-up nudges, “Ask AI” search, and the rule that AI must never add latency or friction* Superhuman’s agentic framework: tools vs. agents, fighting “agent laziness,” deep semantic search over huge inboxes, and pagination strategies to find the real needle in the haystack* How they evaluate OpenAI, Anthropic, Gemini, and open models: canonical queries, end-to-end evals, date reasoning, and Rahul’s infamous “what wood was my table?” test* Infra and cost philosophy: local-first caching, vector search backends, Baseten “box” pricing vs. per-token pricing, and thinking in price-per-trillion-tokens instead of price-per-million* The vision of Superhuman as your AI EA: auto-drafting replies in your voice, scheduling on your behalf, and using your inbox as the ultimate private data source* How the Grammarly + Coda + Superhuman stack could power truly context-aware assistance across email, docs, calendars, contracts, and more* Inside Superhuman’s AI-dev culture: free-for-all tool adoption, tracking AI usage on PRs, and going from ~4 to ~6 PRs per engineer per week* Why Loïc believes everyone should still learn to code, and how AI will amplify great engineers with strong fundamentals while exposing shallow ones even faster—Loïc Houssier* LinkedIn: https://www.linkedin.com/in/houssier/Where to find Latent Space* X: https://x.com/latentspacepodFull Video EpisodeTimestamps00:00:00 Introduction and Loïc's Journey from Nuclear Submarines to Superhuman00:06:40 Docusign Acquisition and the Enterprise Email Stack00:10:26 Superhuman's AI Vision: Your Inbox as the Real AI Agent00:13:20 Ask AI: Agentic Search and the Quality Problem00:18:20 Infrastructure Choices: Model Selection, Base10, and Cost Management00:27:30 Local-First Architecture and the Database Stack00:30:50 Evals, Quality, and the Rahul Wood Table Test00:42:30 The Future EA: Auto-Drafting and Proactive Assistance00:46:40 Grammarly Acquisition and the Contextual Advantage00:38:40 Voice, Video, and the End of Writing00:51:40 Knowledge Graphs: The Hard Problem Nobody Has Solved00:56:40 Competing with OpenAI and the Browser Question01:02:30 AI Coding Tools: From 4 to 6 PRs Per Week01:08:00 Engineering Culture, Hiring, and the Future of Software Development This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe

We, as a human species, like, we started to write because we didn't have, like, enough storage for stories that we were telling to each other. So we had to write to store those stories. Now, like, all the content can be stored in YouTube, in TikTok or whatever. It's like, what's even the need to write? What's the need? Because everything can be vocal. And I see kids now, they don't read article.

They want a TikTok video talking about the article. Being a bit more grounded, what does that mean about, like, the future of the user experience for email and communication? will people still type or will they just talk to emails and they want to hear an email? And this is where it becomes interesting because Rahul as a CEO, maybe next year, he doesn't want to write to you with the new feature. Maybe he wants to talk to you. And then the way you will have received our marketing campaign about the new features, you and your car commuting, listening to Raoul talking about that.

Hey, everyone. Welcome to the Latin Space Podcast. This is Alessio, Fondero Kernel Labs, and I'm joined by Swix, editor of Layden Space. I just realized I have the tough job of always pronouncing names. I know, man, you got a prep. You're on YouTube, name pronunciation.com. Lloyd Cousier, welcome.

Wow. I'm impressed. Did I get it right? I know. You got it right. I'm surprised. Usually I make it a joke about like, yeah, you know what? You can delete it like the way you want and everything.

But like you nailed it. So I'm impressed. Thanks for having me, guys. Yeah, of course. Thanks to coming by. So your CTO is Superhuman Mail, which is the new name for Superhuman. I've been using Superhuman for a long time. I think I was one of Rahul's personal onboarding things back in the day.

And yeah, we're here to talk about all things, AI engineering, but also you have a lot of history in products board, first base, docuSign, and nuclear submarines. Yes, yes. That's kind of like the fun, like icebreaker that I give to people sometimes, like, like to truth and a lie. Like I went into a submarine. And people are like, yeah, no way.

But I did. I did. I spent one year. working around submarines and the trajectory is a bit weird you were engineer and then you were sort of chief of staff on some submarines thing yeah so then you went back to engineering i i started like studying math so i'm a math graduate i was about i was about to do like a PhD in math and applied math in cryptography so like crypto before crypto to some extent it was cool for a moment and then i was like no way i spent like three years of my life on the same topic but in the same lab there was a bunch of

people doing like security, like offensive security type of stuff. And I was like, that's what I want to do. So I was basically an engineer, I was a security researcher in that lab. But I did that in a pretty big corp. So one in Telco and then in the defense industry. And the defense industry, they have this nice kind of like career framework, like you're young, high potentialish between quotes. So they want you to do like different type of jobs and kind of like have a spiral of career

so that you at some point reach to the C level eventually. So they gave me the opportunity to be out of the tech industry for a year. And I went in a harbor and I was there as, I mean, financial controller, process improvement type of person. And basically helping people do a better job, which was interesting because I had no clue. Torpedo system, radar systems, like even like nuclear engine inside a submarine. But still, I had to help people take a step back from what they were doing and everything. And that was really fun. because I came from Paris, came with my tie, and my suit, and my ego.

I was used to drive people through my technical legitimacy on the security space, and all of a sudden I didn't have any technical legitimacy at all. But I still had my ego. So, like, it was a pretty fast ramp up, like, who put my ego in my pocket, and basically drive by questioning people, like, how do that work? Like, help me. Like, I don't get it. And just by questioning, I kind of, like, build a new skill,

which is like getting curious and understanding how people are working and being comfortable, facing people are way smarter than me, knowing better their field, but probably having a way to ask questions to help them, like identifying gaps

or like productivity gaps, for example. So that was cool, but I missed the tech. So I moved back to the tech industry after basically two years. Yeah, what are some of the other maybe highlights or stories you have been told about other experiences? I mean, DocuSign is another product that we all use. Yeah, any other...

No, DocuSign was cool. I mean, DocuSign was cool because it was an acquisition. Open trust in DocuSign, yeah. Yeah. So I was a CTO of a small company in Paris. And we were like a typical, I was a European company at this year. So like very focused on the tech, not very focused on the marketing. And we are trying to, like, we were one of the biggest signature company in Europe.

But it's a very fragmented market. So we were winning France, starting to expand. And Do QSan is coming and like, hey, guys, we need to do a partnership and everything. And pretty soon they understand that European market is tough. And like the technology behind DocuSign is not sufficient, like of standards, like of compliance and everything. So pretty soon they were like with us or against us. But the way they were explaining the value of like, holy cow, like we're not talking the same language. We're doing the same job.

We're selling the same type of software. But like we're talking to CIOs from a technical standpoint. They're talking to head of HR, head of functions, and sell them the value. So pretty fast, it was easy for us to understand like, wow, wow, not the way to set a product, better to partner with them. So they didn't an acquisition, but it's not a full acquisition. It was a security-oriented company, two business line, one doing signature, which is, I would say the one that Anto-Crysan was interested in. The other piece was doing strong authentication. So PKI stuff, SSL certificate.

those type of things. And we were working for the Department of Defense in France. So we had the Ministry of Finance in France, as you're saying, no way, no go. You cannot sell. So we had to do a carve-out, which is like the funniest acquisition type you can do. So you have your team.

You need to divide everything into your team, your systems, your quote source, and all of that. Even your data center, you have to replicate and get rid of like all the shared systems and everything. So we did that for like something like six months, to be able to sell the new Carvout company to the QSai. Crazy. Don't do that.

Are you still involved at all with the French startup ecosystem? I'm curious, like, how you've seen things evolve since then. Yeah, it's pretty interesting. Like, I've seen a change. Now that I'm getting some gray air and I have some experience, like I try to give back to some extent. So I spend more time helping the ecosystem there. But it's funny to see, like, the difference.

Like, when you're here, we live in a small bubble. It's crazy to see how even like other tech scenes are different. So like the great, like just like the great, like to get it done and like to move forward and everything. They have great education when I said there. Sorry. Like we? They? I don't know where am I now.

So great education. Great engineers and all of that. But not the mindset of like creating things. So not a lot of entrepreneur that much. It's changing. We had like some successes in. Europe, and especially in AI, like there's some, some cool stuff happening. But still, like, the way to think about product-led growth, like superhuman nailed it,

but ways to think about, like, the way to structure your organization to scale fast, the level of ambition as well, how to, like, maybe not target France or target Italy to start with, target English and the world from the get-go. And that would be something to think about. So I'm doing that quite some. highly rewarding but it's pretty cool

there's a common question that people have about docket sign that I'm just going to indulge what do all the people do with docian I love it

I you know this is a meme I'm sure inside of do why do you need like so many people do you have signing why do you need 3,000 engineers

it sounds crazy but like you want to make your up you need a different product you need a different team to your local data centers

because of the compliance. You cannot just run your data centers from the US. So you need a local team there. And by the way, the way to do digital signature in Europe, totally different. So like the stack itself is different. So like the way to make a digital signature is different. Not the same standards in the same ways. So you need a dedicated team to maintain that thing.

The same way, some people want to have docusine on-prem. So you need a team building appliance to basically plug and play and like, okay, you have your docisein, applying. There's a doxen box? There's a docu-sand box. Wow. Acquisition made in Tel Aviv at the time. Wonderful people are building, like, security appliance and where, like, you shake the box, the keys disappear.

Like, if someone is, like, stealing your box, no one can sign in your name. Are you kidding me? Oh, my God. No, I mean, some banks. What if there's an earthquake? Yeah. That's a good question. They are mounted, like, on some, like, a earthquake mitigation, I would say, associated to this.

So just that, but like, apply to Fed ramp. Yeah. Dedicated teams, dedicated data centers. And like, oh, and we need, I would say, to have like a docusine run in Canada because that are residency. Oh, we need the same in Australia. Okay, cool. And now you have like something even different. We want Japan as a market.

Oh, but Japan is not signature. It's an ankle. It's kind of like a stamp. So you need a team to understand how Japanese market is thinking about even processing an agreement. Totally different. And then you have like verticalization. like some different verticals and everything. I mean, it's a good business.

It's well run. And like, people are not coasting there. So there's a lot of work. And it's very interesting to see it from the inside because when you see those memes, you're like, yeah, I know. But like, damn, I see people. You're the VPN. So you know, you know, you actually know.

Yeah, yeah. I just wanted to get that. So I hope it's providing some. This episode is not about Dukestine, but we have to ask. No, no, of course. Of course. Totally legit. Yeah, let's talk about superhuman.

So you joined January 2025. Just get people a lay of the line of like super human AI. I think a lot of people that are listening are familiar with the email client. Yeah. I think the AI stuff is generally new. So just maybe you get the canonical definition of what you want to do with AI in superhuman and then we'll kind of take through. The main driver is how you can put AI in the product to accelerate the productivity of people. It's not to just like do AI things and like the sparkles and everything.

We don't care that much about it. Our people are pretty high expectation oriented and they don't want to slow down. So you cannot add latency. You cannot. So everything that we do is done in a way to improve the productivity of people. So AI included. First thing that we started to do is like auto label emails. Like is it a pitch?

Is it marketing? Kind of like typical classification that you could do. And so people can say, okay, everything that is a pitch. I will look at that like on the Friday. So, like, during my days, like typical days, I don't look at it. So, like, that was one of the first things. Summaries, like, you have a long thread. What is this thread about that someone, like, shared with me?

Okay, you have, like, a quick summary. So nothing that is, that was very, like, groundbreaking, but, like, just well-thoughts. Just, like, adding things that make sense at the right time. Another example is now, like, we automatically detect if one of your email requires an answer. And if no answer after two days, popping up, hey, this one needs to be like, I would say, you need to send another email to the person because you didn't have an answer. So that was the first step. Second step was like, you know what?

The draft is already ready. You can just hit send. So it's very subtle, but it's like adding a, oh, damn, shoot. Yes, I wanted to remind people to give me an answer and the draft is already there. Pretty cool. Send. And now we have like more and more of that. Now it's detecting, oh, this is a request for you to ask for your availability.

Oh, you have an executive admin that is doing that for you. Your draft is like, hey, let me see, see the right person. And boom, so that it's ready, it's done. And the typical chatbot, because more and more of the use case we see and people using AI inside superlerman is to query your emails. A good example. As I was a tech people, we receive like a bunch of substack, like a bunch of newsletes. like a bunch of newsletter. I would say some are great.

Sometimes like, the content is May. I probably have like, I don't know, 30, 40 subscription. Because everyone has like something interesting to say at some point. And I'm like saying,

no, I don't read them. I auto archive those. And like every week on the Friday, I just like ask AI, which is the name of the feature. I ask my email. Tell me about like the summary

of the older substack that I received this week. What should I pay attention to? And then I can deep dive, you know, I was at a place where I want to, to pay attention to. So this is always thought in a way to accelerate, I would say, the pace and try to not

be in your way. Hopefully, feel free to ping me. Yeah, it's another case. I would say, I don't know if this is a recent change, but I feel like ASCI, I've started using it a lot more. I've been a superhuman user for many years. And you've had it a while, but somehow this year it kicked up a notch. And I don't know if it's because anything changed the product because I wasn't using it

before or is it just me trying it again? Now, that's a good question. Yeah, that's a good question. I think people are more and more used to the muscle of querying things. Because chatypT. Yeah, yeah. So the general consumer behavior is. Yes, exactly.

So the user experience, people, I mean, now like every single product has a chatbot when you can ask questions. So it's becoming like more and more natural to ask questions compared to managing like a to-do list of emails. And agentic search as well. Like previously, I was like, oh, you have to embed my documents, and then it's just going to retrieve.

And like, that's not what I want. But agentic search where you can actually figure out what do I mean when my question, when it asks, it's like half formed. You expand it and then you actually answer it. It's actually really good. Yeah. And we spend a lot of time on the quality of the answers.

So quality of the answers, and you talk about the agentic framework, but one thing that is. And this is a framework. It's not like your own framework. Yes. I mean, we've done a lot of iteration. And there's a lot of subtleties and like multiple pieces there,

but in multiple different models based on where they're like really good at. But where we spent quite some time lately is like around quality and making sure across different dimensions, but like making sure that we are generally good for typical queries and very optimizing for them. And especially one thing we try to to solve for is agent laziness. So through this chatbot, you can, one of my use cases is I receive a Slack,

And I'm like, hey, Eloy, can you review this document, please? Because whatever, it's a tech, I was a tech strategy document. I need to review the doc. I take the link. I go to Ask AI. And I basically pass and say, hey, find me 15 minutes tomorrow. I need to review this doc. And I don't need typically the agent to say, hey, I found this slot and this lot and this lot, which one do you prefer?

I just asked for 15 minutes. Find it. Do it. I have an admin when I was asking her, like on Slack. Find me 15 minutes. She's not asking me if I need care on the morning, on the afternoon. She's not doing it. So working on this agent laziness because the handoff they were doing to the user is losing time.

So like working on like making things happen faster. We spend a lot of time on this. So that's why you might have felt like that the overall quality is better. Yeah. My old joke was because the way that you trigger it is you actually type it in the search bar. and when I was trying to normally do search, it would sometimes accidentally trigger the Skii. And I was like, my joke is like most of my AI usage

is just accidental because I actually wanted to just search. But then I started just using it more. And then the kind of questions that you ask changes. Yeah, to like find people's phone numbers, stuff like that. It's like, hey, what's... I used it to find my contracts. Because I have so many contracts, right, from all my sponsors and venue things.

Like, yeah. Yeah. Yeah, one of the use case that I, I would say, that blew my mind. I was looking for like, I was at a conference. They shared with me like a PowerPoint link, and it was like six months ago. And I couldn't find the deck because I wanted to reuse some of the content and everything. Couldn't find it for whatever reason. Aski, I'm pretty sure they shared with me like a PowerPoint link or something like this.

Can you find it? So facing the contacts and the link. Like couldn't. Yeah. I save like probably 30 minutes, like searching through my emails. So it's pretty cool. It's too useful. Yes.

Because there's no way you can fit all your email into a context. though. Yeah. Right. No, anything else that's more complicated that? So we have to do some pagination because if you do, like, let's say I'm doing that. Like, oh, I'm pretty sure I had a conference I attended where they shared like a link with me. In my case, I don't do like plenty of conference, but still, someone like Rahul, my CEO is basically doing a conference every three weeks or something.

Not kidding, but the use case. That is his job. And he's fantastic at it. Damn, I'm learning so much from there. But clearly, I would say, depending on the use case, I mean, of course you have more than 40, 30, like even hundreds of emails, and that can semantically be close to your answer.

So you need to go through that. So we had to implement a pagination search. So like semantic search for like the first, I would say, 40 deep search. Not that one. Okay. Next 40, next 40. So I'm kind of like using this agentic loop. And while you don't have find the answer,

continue. and even extended like the semantic search proximity until you find the right one because it might be buried page two of the search page technically. How did you design the tools to get to the agent? Just maybe give people an overview of like the framework, what it looks like. Like how are you structuring these interactions? Is there just one superhuman agent that does everything? No. Do you have separate ones?

We have separated tools clearly. So even an agent, like I would call about tools. So there's a bunch of tools to detect your availability. the tools to understand who are the people you interact with, a tool to write an email. The tool to, like, so every single action is very tool specific. So it's not a magic big tool that can do pretty much everything. It's a set of small tools that are used within the gigantic framework. So like there's a first step that is like, hey, what is the best tool to do this?

Kind of like building a plan. Like for each step, what is the tool and then making the calls? Yeah, I think now the tools versus skills that Anthropic talk. about is like the hottest thing of how much you want to put and there's like the MCPs discussion. I'm curious how you evaluate the tools to like when you build them. It's like how do you think about how to name them? Like how to give the description? It's like how much work have you had to do to nail it?

I don't think we spend that much time into and again like I will defer to my three engineers working on it, which is interesting. We can talk about like the amount of people you need to work on those stacks when you want to be serious. And I have fantastic people. So I feel blessed. And most of the time was trying, like, the different agentic framework, trying to understand the different models, the ones that are solving which type of problems,

because every single model is good for something. Sonet was really great for, like, agent head off. Like the laziness was really great. Open-E-E-I version of it was not that good. Now we have Jimini coming in the room, like, in the room, like, last week. Like, that one is cool as well. So I think we are, I guess everyone has appealed like a way to form one switch easily from one water to the other. So like model routers.

Everyone has like an LLM proxy to some extent and like an agent proxy to implement different stuff, which is becoming interesting because the way to tweak them and tune them is different. So it's still easy to switch from one, I think, from one, I think, framework to the other. But at some point, I think it would be harder and harder and like the stickiness of them will be tricky. But to answer your question, like, we didn't, didn't. spent that much time on the tools themselves, I believe. How do you think about evils? Like, are you evalling, like, one email draft at a time?

Are you evalling a longer workflow? It's like, just run us through, like, yeah, when you're testing Gemini, like, how do you decide what it's good at, what it's not good at? Well, it's like the email structure. At first, we had a relatively naive approach, query answer, query answer, and having, like, a set of queries. We, over time, evolved into, like, thinking more about, like, the different dimensions that we want to target.

Agent Endoff is a very typical type of problem space that you want to make sure you select the right model for. So typically getting a bunch of queries targeting hard handoff that we've identified by through the footing or whatever, but trying to target a set of what we call canonical, I would say queries along that dimension of, I would say, that specific problem space of agent handoff. But there's more, like, there's the deep search, like shit ton of emails and you want to find that needle in the hashtag.

That's a different type of category. So you need to have canonical queries that are like targeting that type of dimension because every single user will have their own way to question their own data set. And we cannot replicate every single data set of people. The good thing is we have a bunch of users like Raho like myself. We receive like a shit ton of emails. Not on my French, by the way. I don't know if it's okay for the show. But he receives probably like $500 to a thousand email a day.

He's still part of the onboarding. It's like, I'll send an email to Rahul and he will reply. I'm sure it's not actually him. Sometimes it's him. He's reading like pretty much everything. I don't know how he's doing it. But he is really, really paying attention, especially at the tone and why something is like going sideways and everything. He really associate the brand and tone of like the people talking the company with himself,

which is kind of like bringing us to the next level as well. So thinking about all those dimensions is really key. So like even if you have like an eval tool, like the way you structure your different queries to target those dimensions is important. And then we have those specific queries, like the raw queries, typically. The one we joke about and the one that was one of the first we used as a way to calibrate our quality was weird stories. but he did some, like five years ago, some refurbishing in his house, and he had this table, specific type of wood,

and he was discussing this with the contractor. And he wanted to have Haskii find that email and the type of wood that was discussed in the thread with that guy five years ago. And until we nailed that query, he was not satisfied with the deep search approach. And this is where we were like, oh, damn. Okay, so that's a different set. But we're also talking about dates. Like another dimension is dates.

What is last quarter compared to today and everything? Large language models are not really good with dates. So like, how do you manage that? So these specific queries for that. So we're like, oh, okay, so there's the dimensions that we need to care of. So now we structure the all evas. And as you are asking, end to where? What is the query?

Whatever happens there? There's like an answer. Was there like a good agent end off? I would say date where they're nailed or not. and, et cetera, et cetera, et cetera. So it's pretty intensive in terms of brain power put in the quality. Again, because superhuman is a high perceived quality type of product. So we had to invest that amount of time there.

Yeah, high real quality. It's not just perceived. No, but I think this is important because what is quality? I don't know. The feeling. Like if I buy a car, like that is a Toyota, it's good quality. And I get the quality for my box. if I buy an Audi or Porsche, I expect a different grade.

So maybe it's great. Like the grade is different, and it's high grade, but high expectations, so a high amount of time spent on quality. Yeah. In PMing, there's this concept of the high expectations user. And Rahul was one example of those. And I just wondering, like, who are the most outlier extreme people? How are they using AI in their email?

You know, just in general, like, the most extresteenable. stream examples that you've come across, obviously, because that's how you work. Oh, that's a good question. For example, you had how much time do I spend in Waymo's last month, right? Which is basically turns your email into an accounting system because it's kind of a source of truth. I don't know if I would do that in superhuman. Is it reliable? It is reliable.

Wow. And when you think about the amount of work, and we're working right now with Anthropic to basically do, like building on the fly, small, kind of a kikonot of lambdas that will bid the code to do the aggregation. This is an easy example. This is like a code execution thing. Yes, it's a code execution piece. But like this one is a relatively simple because you just, just have to have like the agent extract from the email.

So select the emails from Waymo, from the waymo, receive, like extract the time, like the duration of the trip and then do the aggregation. But that's not easy. Like that aggregation is not easy. And LLMs are not good at math. So like that there was some support about it. And right now we're discussing about like, like, extent. standing this approach to more. Are you operating on the email file itself, or is there a fundamental, is it like a role in a database and you're just writing a SQL query?

No, the aggregation is. So we don't extract that data on the, so when we ingest the data. So we ingest the data. Yeah. We ingest the data. So we rely on Gmail and outlook, of course, because they are doing like some great stuff that we don't want to do, that detection. And superhuman will never do it. And probably.

Probably. Probably. Probably. Which is being a IMAP server. Exactly. Do I want to do that? Probably not. Probably not.

You know, hey, hey mail, did it? Yeah. But like, is it something where we want to spend time? Is it valuable for our end users? Really? Not sure. They live in a code system. They will live in a different company.

Outlook. Yeah. So, like, they have outlook and they have Gmail. It's already there. So, like, if we can just plug and make that better, I mean, it's, it's good. I mean, in some case, Superhuman was the original rapper company. If you think about GPT wrappers, this is the Gmail rapper, the Gmail rapper. At first it was LinkedIn rapper and not Gmail rapper.

I think more for it than Gmail itself. It's very true. It's very true. That said, you can question what is like an SMTP server for real? It's a server that conforms to a spec with some database. Maybe not even. Maybe not even. Maybe not even.

I mean, they're doing like way more stuff. Like they have like crazy, like especially GCP. Gmail, like the search capability is, of course, like, I would say, crazy good and all of that. To do what you do, you need a server-side clone of my Gmail, and then you need also a local cache. We need local cash. We work offline. That was one of the things that we did, as initially, beside the UX, beside the speed. We have everything local. One of the reason is we want to be fast and under, like, every interaction should be under 100 milliseconds. Yeah. I mean, with network, you cannot.

You just can't. So everything needs to be local. So yes, so we have like a copy of emails local on device and works in the enterprise world because interestingly for mobile it's used to be Realm. Yeah, RealmDB. Yeah, yeah. Is it Facebook tech? Mongo. Mongo.

It has been acquired by Mongo. But now it's like somewhat unsettled. So we need to find a different way to do things now. It might be SQLite. But yeah. So on device stored, but that was like the old search where we had basically like a database with rows of the emails. But everything that is AI, like we have all the Vend headings and all of that. So we have a hybrid search and we use, I don't know if we can name brands, but we use the TurboPuffer on the backend to store like five years of this story.

I think it's relatively public with their customer list. So I don't know. No, yeah. We'll let the API department. They talked about it anyway. But it's a, I mean, stable infrastructure. They do things pretty well. It's fast.

I'll briefly comment that I know any number of local first database companies that we would love to work with you. If you're saying that you're on the market for a realm replacement, they will come and talk to you. I mean, I'm more than happy. I'm more than happy. That's my AI, my mobile team. Like, they're really looking for like something.

They would love nothing more than to be superhuman's database. Okay. I want to just like focus on the AI side, right? Sure. So people want to know where is their inference running. What are you sending over? what can the providers see? It depends.

It depends. It depends on the use case, depends on the type of model we want to use. So there's some stuff we run on inference company with Open Models. There's some stuff that we run with Open AI, we're Anthropic. It's pretty diverse. It changed because also based on the quality of the models. We are a GCP shop.

So lots of credits for Gemini? Yes. So we have an incentive too, probably like I spend some dollars there. I mean, it's nice that they're also a leading model anyway. So you're not actually compromising. They're doing some pretty good stuff there. But we use Best 10 to run some, I would say, some Lama, some birth model for classification. There's, we're doing probably some discovery discussion with like some YC companies about like a model on device as well.

Because. Yeah, so work offline. Yes. And interestingly, those companies there, started to do on device, mostly for cost reduction. That was their pitch. We'll reduce your cost. I mean, we don't care that much. Our people, our users, they want quality.

And they are okay to pay for that quality. But we want to solve for offline. Like, if you're offline, semantic search doesn't work as well. So we are discussing with a... What are your design constraints for offline inference? For example, right? Like, deep seek V3.3 would be like, 600 billion parameters.

I don't think you want to take out 600 gigs. And people are somewhat complaining about like our footprint. Yeah. On the device. It's probably like two gigs already. Both in memory and both on device because we store local emails. Like we store like when you install superhuman, we don't know the last 30 days of emails so that we can do search when you're offline at least for the last 30 days.

But we keep that history. So it's starting at 30 days. And if you're like a customer for like two years, technically we optimize for two years of email in your device. So that's interesting. On the local model, any thoughts on like every app is going to have its own model versus you're going to have a device model that people run? I mean, it's a lot of space. What would you prefer? I'm curious.

Would you rather have the user just take care of the inference and rely on that or do you want to own the whole experience? I mean, Superman will want to own the food experience. Like, we're pretty picky in the way things are, I would say, happening. But at the same time, like, if we talk about mobile, you want the mobile experience feel like your device. So we are basically not doing a React Native. We are doing Swift. We are doing Kotlin because we want the app to feel like the user experience generally on iOS or on Android. But for the models, that's a good question.

I would love the device provider to be better. I mean, we can question like local devices. Like iOS has done some work there, but it was underwhelming so far. They're still working at it. And that's why we have like a YC companies that are spending time there and doing some cool stuff. Yeah, amazing. Interesting question on base 10. They are a very different cloud inference provider for open models compared to, let's say, the fireworks and the together AIs.

The general pitch is that they don't charge by token. They charge by box effectively. Anything else that's interesting working with them versus the other inference providers that you buy? They're easy to work with. Yeah. I mean, that's when you're a startup, you want to move fast. They're really easy to work with.

So the priority is like what cost, speed? For us, it's quality. So it's quality and speed. It's all the same quality. We would always start with the highest and more expensive. model to get the right quality. And when the quality is nailed, then we can spend time trying to optimize. Right.

But all these providers, Base 10 fireworks together, all these, they all have the same access to the same models. Fair. So unless they quantize heavily, which all of them say they don't. So in that case, like the fact that it's a box, you control your cost way better. Yeah, yeah. So it's like fixed capacity. It's fixed capacity.

So you know when I discuss with my CFO, like, yeah, when it's token based, it's like The exercise is way more trying to understand, like, I would say, the adoption and all of that. But that's serverless. That's serverless. Sure. Use it case, serverless is scales up, skills down? Fair, but like the cost control is becoming like a thing. It was the thing before the acquisition.

Now that we are part of a bigger umbrella, like understanding your cost structure and like being able to make projection that are closer to the reality is more important. Like all pre-IPO-ish companies, you want to really. understand where you will be like in three months, six months from a cost standpoint. So based on for that, it's pretty cool because you have more latitude to stay within the bracket of like a box. Yeah. I was thinking about this. You know, a lot of people think about cost in terms of dollars per million tokens.

Sure. And I think that that is actually amateur thinking. There's only the kind of pricing you care about if you're a solo developer. But once you're in a large scale like you guys, And it's also something I learned about a cognition. You should actually cost here about price per trillion tokens. Because we spend multiple trillions per month. And when you unlock that scale, you unlock different ways to spend.

That's not a serverless token-based pricing. So basically, I think base 10 makes a lot of sense on a price per trillion. Yeah, I didn't look at it that way. It's pretty interesting. But no, no, that's fair. And I mean, we built like so many different models, not trying to understand like the cost per million of tokens. And then you have to infer, like, what is the average number of tokens?

Because we treat every single email. That's really short emails, very long emails. It's like you have to understand your data. Like, what is the median and all of that to make your protection. And it's always, there's always some magic. The reality is like you don't have the time to, I mean, I'm an advocate or like, let's move fast. And if it's successful, it's great, even if it's expensive. So rather than trying to optimize the cost too early, like just go.

with something that you control and fast, and you'll have time. I mean, it's a good problem to have. Success is a good problem. When do you think it's going to break from like a cost perspective? Say you were to like draft every single email that I get. I'm sure you will lose money on the 40 bucks a month.

Yes and no. I think that it's a matter of like how more productive we make you. Like we have some customers that told us like initially we were talking about like the different models and everything. I'm like, take the better model. Like I'm ready to pay like 200 bucks. month, but like get the made, the best model. Like, I don't want half crap because it's less expensive.

So, like, always give me the best. Because these are all, like, high value. CEO doesn't these. I mean, one hour of their time is worth, like, 10 times the, the amount of the subscription. So why isn't there a $200 a month? That's a good question. Yeah.

I'm not in charge of the pricing and packaging. Okay. Maybe, maybe once in an example would be like, well, what's one thing that you would like to do that you cannot do with today's models, even though you're try pushing quality. Then your customers are telling you, actually we really want this or maybe Raou was telling you that he really wants this.

I don't know. Yeah. I don't know. I think we have the means. We have the means to do like pretty much everything that you want to do. Like it's a matter of executing and doing it right. The way I'll put it is like if you can articulate what you cannot do today that you think you should be able to do and your customers would pay you for it,

the model less will make it happen. But the problem that you have and the problem that I have with Kong, is we cannot articulate what it is. We will know if it's better, but only once it exists. No, that's a good framing. And the other piece that I think it's pretty tricky is that there's a transformation that is happening in the user experience.

Like even the way we're thinking about the user interface right now, it's totally switching. Like the way we think about emails right now, it's still like some sort of like a to-do list. It's a table to some extent with rows. What would it be like in a year? because people will be more and more interacting with their systems through a conversational aspect. Like I see my kids.

My kids, they don't type on their phone, they talk. I mean, all my kids have three kids, all they talk with their phones, working college and middle school. Okay. On WhatsApp? WhatsApp, because they're European and they need to talk with the family. The reality is like Snapchat, it's like a TikTok, like whatever, like Instagram. like they communicate over Instagram. Like, that's not an image tool, like or something.

I feel like a boomer. Yeah, I am. I am. But what is interesting is that they, and we can debate about like, but like we as a human species, like we started to write because we didn't have like enough storage for stories that we were telling to each other. So we had to write to store those stories. Now like all the content can be stored in YouTube, in TikTok or whatever.

It's like what's even the need to write? What's the need? because everything can be vocal. And I see kids now, everything is vocal. They don't read article. They want a TikTok video talking about the article. So coming back, and I'm sorry, like I'm getting like very high here, but being a bit more grounded,

what does that mean about the future of the user experience for email and communication? Will people still type or will they just talk to emails and they want to hear an email? And this is where it becomes interesting because Rahul as a CEO, maybe next year, He doesn't want to write to you with the new feature. Maybe he wants to talk to you. And then the way you will have received our marketing campaign about the new features will be discussed to you or talk to you with his voice,

not just voice and tone in terms of like writing, but like you are really like you in your car commuting, listening to Raoul talking about that. So coming back to what cannot be done right now, I think like the main problem is like nailing the new user experience. I mean, open AI now you can do something. stuff with emails. They're trying to do some stuff there. Like all those chatbot, they try to be like basically the new OS to some extent. So how do you interact with those new apps? So what is an app even in this new world? So that's what is like really interesting. And that's why I'm glad to work with Rahul because the guy is so freaking visionary. And if there's one company to nail it, there's not a lot. And I believe like a superhuman is one of them. Yeah, I think the inbox is like

the ultimate private data source. I feel like even when I see all these companies that are like, you know, talk to like your AI clone to get advice or like, you know, things like that, I feel like so many times, man, I'm just writing the same thing over and over. Like, you know, how many founders email me asking about help for XYZ task? And like the answer is almost always the same, you know? And like there should be a way almost for superhuman to like be the advisor on my behalf in a way.

is that you should be able to predict what I will respond to this email. It's called auto draft for respond. We're testing internally because like there's especially sorry to catch you up. But like same for me. Like how many companies are reaching out to me to pitch whatever like AI frameworks or like AI tooling or like whatever? And my answer is like although I don't answer because I receive like on loads of them. I'm saying like, thank you. Don't have the time and everything that suits cool.

But like because I want to be polite like right now like like. it's automatically generated for me because they learn that I usually don't care. Yeah. And that's my answer. Or if it's someone that is pitching me for like, hey, I want to work with you guys and everything. Like someone that is applying. My answer is usually, oh, please reach out to HR. I'm seeing HR and everything.

So now we are, I was able to understand how you reply typically. But it's always like if it's covering only 80% of your use cases and you need to discard 20%. Where is like the cost-benefit value? Is it annoying? To have like 20% where you like this car? I want to write it myself. Is it good? Like what is the limit?

90-10, 80-20? I think it's like AI plus the snippets that you have. I think that's kind of like, like I have snippets for a bunch of things like vendors. I have just like super long snippet. Thank you so much for reaching out about your company. Sounds like a great product. We're not currently in the mark, blah, blah, blah, blah. And then the response is like, thank you so much for your thoughtful.

response and I'm like, great, get it out of the way. But I feel like if you could use that plus AI to do the small kind of like last mile thing, I think that would be enough. You don't really need a GI. I'm excited for it. Q1, Q1, Q2, something like this. I pay 200 bucks a month to open AI to one tropic. Like, I'll give you 200 bucks a month if you like make me not write the same thing over and over deal. I think more generally what he's trying to get at and what superhuman is starting from a very good basis, but not there yet, it's kind of like AI EA. I don't know if this comes up a lot.

Why have people I work with who do read my emails and respond for me. Yeah. And they have memory and they know my normal preferences. They have human judgment, which LMs don't have. Is that something that you would want to build or do you think what you want to leave to others? That's a goal. When we kick off really like the revamp of our AI world and what AI means for like superhuman, Rao did a pretty good pitch on it.

And there was like a pretty nice video. I think it was in March for the launch of like the new AI. That's the vision. Like the vision is like you have an EA. And most of the people are using superhuman. Seasuit, founders and all of that. So pretty fast they need someone to help them with their emails. And we want to do like most of that job.

So we're getting there. We're getting there. But that's a goal. That's a goal. Like the first thing like answering your availability. Right. Now we can do it. I mean, right now, it's in beta.

But right now, my emails, like internally, when someone is asking, like, hello, can we meet next week for lunch? Automatically, I will have, like, three slots proposed in a draft, and I can just, like, send a draft that is prepared for me. Yeah. It's still up to you to decide whether or not you want to send a draft. That's the thing. I don't want to be involved.

And this is where your EIA will always be better than an LLM, because she knows the type of people you are okay to have lunch with, Or maybe they have the context because, oh. Yeah, sometimes you're busy, but you're like, oh, VIP, I will move this. Exactly. You know what I mean? I get into it. And your calendar is not going to know.

I mean, we're getting closer because we know how much time you're interacting with that person. But like how much time you interacted. It doesn't mean that maybe last week you had like a bad discussion with them. And now you're not friends anymore for whatever reason. But your EA would know. So there will be like always limitation to this. But and that's why we want two people to always be in the loop. And maybe it's your EA that is in the loop.

It's so helpful when I'm not in the loop. Yeah, we can batch it and like I have my one-to-day call with the EA. But yeah, obviously that will happen. You know, some ways that other people are pursuing this, like Notions trying to go after it, right? They have Notion Mail, Notion Calendar, then obviously they really care about AI. Some other people are doing this interesting thing where they buy an EA company, like a company that already does virtual assistance. And then just monitor what they do and then just, first of all, Superhuman can provide me an EA. that is a human

and then it slowly replaced parts of it with AI. I'm curious what you think about that. That's a more aggressive approach. If you really want to...

I mean, that's probably the best way to understand how and he is working and like the type of work that they're doing and everything. Yeah. I mean, that's intense.

That's intense. But like, sure. You have the money. And you pretty fast understand what are the type of workflows you want to automate first. So like having that data would be like

pretty interesting. One of his portfolio the companies they bought a legal for, yeah. Do you think, do you think that's an accurate description or am I glorifying it too much? No, it's an accurate description. It's like it just behaves as a law firm, though. Right. Just treat it as a law firm and then internally start to optimize.

I mean, you have now so many customers that it might be. You might need a lot of EAs too to do it for everybody. But I'm curious, I think like the, yeah, the memory is kind of like the killer feature of the EA. It's like understanding real time. I'm curious, like, now that you're, like, within superhuman, the company, you know, super human mail? Yep. Do you feel like there's, like, a lot of advantages of being email plus documents, plus being embedded in everything? Like, do you feel like that helps closing some of these gaps?

Yeah. So, for example, like Coda is an interesting, I was a piece of software. So Coda is like an ocean equivalent. Yeah, we used it at Amazon. Yep. It's a pretty good one. And a lot of, like, enterprise companies start to, like, use Coda more and more because of the flexibility and everything.

And CODA has this concept of like CODA packs, which is integrations, glorified integrations, if I would, I can say this in this way, but they're ingesting the data. So like the data is there. So like every time you have CODA, so we have technically an injection pipeline

that can aggregate all the knowledge about you in the company, which is great. And now if you add Gramaly, Gramaly is ubiquitous. Our users of Gramaly, Gramaly knows that you're in Google. Like, Grammarly knows that you're, I would say, crafting a post on LinkedIn. Gramalys knows, technically they can know.

Doesn't mean that they use the data, but they're everywhere. So like when you have this, I'm everywhere. Oh, you're getting into your email. But I know that you were currently like on Jira with that context. So all of a sudden I can pop up like some of the context. I know that you're writing to that person. Oh, it's about this. I can expand and like augment.

your email because I know where you were coming from. So the data will be there through code. Gram only knows basically where you, I'll say, you're switching from Google Doc to Salesforce to LinkedIn, and now you're writing an email. So we have this augmented context even more. So like much more precise compared to something like chat GPT, for example.

They don't know where you are because you're switching windows. You're coming from to, I would say, to GPT, from Salesforce to chat GPT. They don't know where you were. They wait for you to pass the content, to get the context. If you're grammarly, I know where you're coming from. So when everything will be converged, and we've been acquired only like three months ago, but when everything will be converged from a contextualization standpoint and knowledge standpoint,

we know way more. So we'll be like way more accurate in the way to help you. Maybe predicting fourth acquisition, but wouldn't it make sense to have your own browser? That's a good question. I think there's much more to be done on the productivity. space before like I would say solving a browser and everyone is trying to do a browser. Yeah, Elassian, perplexity, opening eye. I'm still sad that Arc is not like getting into development anymore because of Dia.

But Dia has been stopped. They're rebuilding Arc in Dia. Yeah, but like it's, it feels like it feels very unstable now. So like more and more people are basically saying like, okay, let's go back to Firefox. I mean, more and more people are doing that because like there's so many browser. Like you're like, you want to wait for the war to be done. And to have like the clear wing. No, no, no, no, no, no.

I disagree. I disagree. You should go all in. What are you using? I use Atlas. Yeah, I'm also Atlas now. Oh, interesting. I'm still on Arc.

It doesn't have profiles still. Yeah. That's the biggest issue. So I, based on the different emails I have, logins I have, I switch between Atlas and Chrome and Arc. Interesting. Yeah. Yeah.

Yeah, my personal one is on Chrome. I'm just saying like, well, okay, if that context matters to you, right, if Koda and all those things, then grab a million others, you might as well have your browser. This is the season of no one no one will get upset at you for saying oh we have a browser like it will be like yeah it makes sense. Oh it would be like oh no one more.

But it's the superhuman one and that's a good brand. That's interesting. I foresee like a like browser to disappear completely. Okay, that's the title. I mean my main like central I would say piece of software that I use in my productivity tool is recast. Yeah. I mean I'm a Mac user.

So I use Recast. For the people that don't know Recast, it's basically like a way better spotlight on Mac. And I don't need bookmarks in my browser anymore. What is doing a browser besides providing you a view on the website? Nothing. So it's just like, so even like to some extent Recast should be like just a web view. Because what I do with the Recast is like...

Then you're turning recast into a browser. Is that a browser if it's just rendering HTML? Yeah. Okay. So, like, everything is browser. So, yeah, if it's only like a rendering HTML. What else you want? You want JavaScript?

I don't know. I don't know. Local storage? You want local storage? Extensions. Like, you need a browser, like, to have, like, your local extension. But to have, like, a local storage that is, like, pretty massive, like superhuman. But, I mean, what's left?

If you were, like, everything that was making a browser, a browser before, which was, like, bookmarks, like, basically the last history that you had, maybe, like, cookies and like what's if you get rid of that it's just a view a web view to some extent yeah it's a clean application platform with that open app store you know that there's a mark and jason line of well the operating system is just a poorly debug set of device of device drivers for the browser because the browser is the actual application interface um from the person that made the browser yeah yeah i i think the browser would be like more and more thing I believe they would be like thinner and thinner, but they will disappear.

Or they would be like just embedded in the OS eventually. Yeah. So one more technical sort of thing and then we can go to sort of organizational things. You mentioned understanding the person, you know, part of memory is just like the knowledge graph. And one part of knowledge graph that really matters is the entities that I deal with, right? Like I deal with him for four years and we have that context. and basically what exists today in superhuman and maybe what is possible in the future.

For example, do you use a graph database or something like that? Not yet. And incentroism because you were mentioning what's missing right now. I think that these knowledge graph, like, oriented database, I'm not there yet to some extent. But have you actually tried or are you just saying that? No, we didn't try. Yeah, that's the thing.

It's not fair to say they're not there yet if you haven't tried. Correct. Correct. But even from a taxonomist, point when you think about those entities, what are those? If you're verticalized... People, companies. Yes.

But then you start talking about projects, but is the project? Is it a task? Is it an initiative? Is it a hierarchical aspect to those? How deep is the tree? These are all the other questions. I think it's very... Like, you know, superhuman's history is reportive where like the person is like the core of

the universe. No, no, but there's some obvious entities. But like if you think, if you want things to be really personalized, these entities are like very, very subjective. Like I'm a user of Obsidian. Yeah. So I'm a note-taking nerd. And for the people that use no obsidian, it's another local first app. It's another like local first app in which you build your own workflows and where you will basically through templates, define your own entities that makes sense for you.

And there's no two like graph that is similar. Even if you're using the Note app, say, for the same thing. So trying to infer like a generic knowledge graph that can be reused with like dedicated entities, people, task, project and everything. It's harder than it seems. Interestingly, like we were thinking about it when I was at Product Board. Product Board, we have like the roadmaps of like so many tools. Based on that, you can probably infer some taxonomy about. what is a SaaS product.

But even trying to generalize this into like a tree that can be repeatable for people, it's hard. There's some common stuff, authentication, authorization, billing, user management, dashboards, whatever. Every SaaS company has this. But then when you come, you enter like the domain of the company, totally different because their features, their surface area is very different. So like even they are trying to form the knowledge that you have.

have abstract the entities that will be the same for everyone is not easy. So it means that then for each user, you need to have an unoptimized graph that is like subjective and dependent on the people. So you need to build the graph based on the like just the data. And you don't, you don't have like a real way to, to optimize it. But you're fair. Like you're right. We didn't try.

But also because. Many people have failed. It's fine. And I don't even foresee a path where that can be surfaced into more productivity gain. At the end of the day, what is the problem you're trying to solve? It's super nice from a technology standpoint and even like a thinking process standpoint. What is like the ultimate data model for productivity nerd and all that? But what are you improving from an experienced standpoint?

Is it like the accuracy of your draft that I'm playing for you? I want my AIEA to remember everything. I've talked, everything I've done, everything I talk to, everyone, every conversation I've had, you know. Yeah, but then it's Jarvis and it's like almost AGI, to some extent. You have the context to anyone else else. Yeah, but like the amount of compute and the amount, because you need to recompute like your graph every time you receive new stuff and everything.

So it's an interesting space. I think so to your point, we probably as an endpoint solution, we probably won't be the one solving for that. I think that there's like companies that should focus on this and be like, hey, I'm the engine that you will ingest everything that you're doing and we build the graph and the graph would be like the best graph ever and it will be like for each account or each tenant will build a graph for you. That would be great.

But is it something for JobUpuffer? Is it something for like those vector database companies to solve for? Maybe. So for what it's worth, I'm actually dating someone who's doing upside and they're mining emails for the CR, basically like CRM. population and building a knowledge graph from emails. Interesting. So basically, they're happy that you're not doing it.

I'd love to have an intro. Because obviously, if you do it, then you are a very serious competitor. No, but I think it's not easy. Yeah. So I would love to discuss. Sure. I think we would be probably more a consumer of the outcome, rather than the builder of that layer. Yeah.

I think the other big consumer, obviously, would be Open AI. Of course. They clearly want to eat everything inside. chatGBT. I mean, this is a cool exit strategy for such a company. For them, yeah. I mean, like, do you want to build an superhuman app inside of chatGBT? I feel like the answer is no, right?

Oh, the answer is like chatypT or like open AI and superhuman are competitors. Okay. Like this is what we fight against to some extent. We have a different approach, I think, but especially this ubiquitous grammarly presence. We are everywhere and everything. I think we want to be more. proactive because where you work, we can be more proactive compared to chat GPT that is waiting for you to do things to help you do the thing.

So there's reactive versus proactive. I think we're more on the proactive side. But that's the competition. Like just, I would say, fun notes, but like when Rahul is questioning the quality of our, I would say, AI queries on superhuman, is comparing us to Gemini. He's comparing us to open AI. So that's the competition we're fighting against. Yeah. I mean, and speaking of which Gemini, the chat app obviously has privileged access to all of Google.

So they can also say to us. And like the search engine is crazy good. Break them up. Raul, break them up. Awesome. On a more broader side, so you mentioned you only have three people working on AI. What's kind of like the coding AI adoption at Superhuman on the engineering team? Yeah. Interestingly, like our path was, so we started to really think about it like in Q1. a bunch of people using some stuff and everything. We didn't have any data, just anecdotal feedback and all of that.

The first thing we've done is cut the red tape. Like, hey, folks, three, four. I will approve the budget like in one hour. You can try anything you want and deal with the security team, 24 hours turn around to get things approved from a security standpoint because you don't want to do some crazy things. Huge Q1 was like everyone was trying everything. It was really interesting to see how things were working super well on the front end, a bit less on the back end.

We're a go shop on the back end. And everyone working on iOS and Swift, like, eh, not that good at the time. But like a huge adoption in terms of tooling. Also, like on the product side, a lot of like V0. For an XHS? No, VZO, VZO is kind of like a bold vote. Yeah. Because they build next year as sites, right?

Or apps? Yes. We just use it for, we just use it for like a prototyping. To be like as close because we have a founder that is very picky and wants to review the design and like a design on Figma is great. But like when you can click into like real stuff, it's so much better. And Figma is not there just yet.

Figma has Figma. We interviewed it. Sure. It's a, sure. It's getting better. It's getting better. But like as a PM, they, use V-0 or whatever like a tooling like this because it's not lovable.

Superman is like V-0. V-0 is a standard. And again, it was like free for all. Try whatever you want. Free market, right? So free market. And free market V-0-1. Always winning.

It's still a free market. Q2 was more about, okay, let's try to understand where this is working, where this is not working. So compile a huge list of wins and an area where like, ah, to do this, not good. to onboard in a new, I would say code area. Amazing. I used to spend like a full day to understand all the entry point, the dependencies on the code stack that I didn't know.

And now I need like 30 minutes with a code code. And I understand how things are working. Even for me, like I'm not in the code anymore. But like instead of like asking my engineers, like, how are we managing like the refresh tokens with Gmail? Like now I just like cloud code and I'm using warp. Warp. Warp.

Warp. Warp is good. But anyway. Warp, clodod. like how this city is working. And boom, boom, boom, boom. I'm probably like the links to the right file, explaining you like the high level concept and everything.

And I don't waste my engineers time to just answer a question. So pretty cool. So that was Q2. And we started in measuring. So every PR, we have to put a label. I use AI or I didn't use AI. And if I used AI, it was productive or it was not. So trying to understand the lay of the land.

Roughly said, I think we have like 80% of people. that are really flagging the PR. Out of that 80%, probably 90% of AI usage. So it's all declarative. We're not like plugging any tool to measure like the real number of tokens

and everything. And out of those 90% again, 90% of positive impact. But it's not always in the code. It might be like just the discovery, understanding like the layer of the land or like stuff like this. So 81%?

So technically like, yes, it's like a 90% 90 of 80, but by inference, I would like, if I caricature, I would say 80% of usage and happy usage. So like roughly 80% of lines of code written in superhuman. It's not the line of code. Probably more than that. Yeah, because it's the discovery.

Like most of the time you spend is not writing code. It's like trying to understand what you need to solve for. And this is the part that has been reduced in terms of real KPI. And it's AI is not only the only reason why we have accelerated. But in Q1, we were roughly said at four PR per engineer per week. In Q1, Q2, we were closer to five PR per engineer per week. And Q3 were closer to six. So the global throughput, and again, PR per engineer per week, we can debate.

But that's a throughput measure. And it increased quite a lot. But again, AI is only a piece of it. Technical strategy, clarity of what you want to do, organization. There's a lot necessary to that. So we feel pretty good. One question that a lot of like the AI leadership people I talk to have is like, am I supposed to like ask more of my engineering team now? Like am I supposed to like, you know, hire less people.

Should we ship more as a company? I think most the thing about AI is like you can do a lot more, but most companies are not built to do a lot more. Like, you know, especially like if you ship a hundred more features, you don't really have marketing to market, a hundred more features. features, you don't have support to learn 100 more features. Like, how do you think about structuring teams and, like, the expectations of it? That's interesting because superhuman historically was very lean in terms of organization. So, like, superhuman mail, like, we have 50. 50, crazy.

50 engineers. And your user base is roughly a million? Yeah, like, less than that. Okay. Like paying users, probably 100,000, something like this. So it's still like a. relatively small. He's still supporting a lot.

But it's, yeah. So it's, I was a small team, pretty senior. And the average tenure is probably four years. So like long tenure, fully remote as well, which is interesting. So my AI team is distributed between Patagonia and Canada. So access to a different pool of, I would say, right people, not trying to compete in the bay. Because people want to go to Anthropic. They want to go to open AI.

And like those guys, like, they have like, They pay too much money. Yeah. I mean, it's not the same, I was a competition. So we find the people where they are and people that don't want to move to the bay and all of that. And there's some great people there. Anyway, long story short, relatively small teams. And we increase the capacity.

We try to not move too fast because we're qualitative. Because it's kind of like a vicious circle. Oh, we can do more. Let's do more. But like all of a sudden, like the number of like bucks coming in. Right. Yeah. It's also growing and everything.

So we try to be. Now we're working on the Gramerly slash the new superhuman. So there's also like an incentive to be like to invest a bit more because it's a product that is working. And Shishir is really willing to implement a model that is called like the compound startup. We're still a startup within Gramerley. So we have our own P&L. We have still like Raoul as a founder. The only difference between now and before is like our board is Shishia on the exact team at Gramerley.

at grammarly or like superhuman. But we want more people. We want, I would say superhuman to have like more reach and like to do a bit more. So now we are kind of like scaling that and we're adding like a more capacity. So AI is helping, of course. But it's also helping like for the onboarding. It's helping for like a lot of that. But we're adding some capacity.

Yeah. Yeah. I think like, you know, the mainstream maybe pushback on it. It's like, hey, like you used to pay me X to do four PRs a week. So am I getting paid 50% more than? and then I just shift six PRs a week. I think that's the thing that that's why there's a lot of pushback around AI as well from people. It's like, hey, look, you know, I'm using this and you're getting more out of it, but I'm not getting more out of it.

I think it's like the usual like, you know, a strong. I would say disagree with that statement. Yeah. I disagree too. I mean, I'm saying like when you listen to people outside of our bubble, yeah, there's like a lot of like this discussion around, you know, where the value is accruing. So basically, if you only look at it as you're paying for output. was the previous payment wrong or was the current payment wrong? One of them is wrong.

Exactly. No, no, no, that's an interesting point. The way I see it is like engineers are well paid. Like we are like a very fortunate. I was a part of the population. Our salaries are probably, I was pretty good and part of like the top, whatever, like five percent in the country or like even in the world. I think that when we talk about like the Maslow pyramid,

like engineers at some point when they're like pretty senior, They don't rush for like 10 more K or 20 more K or like, I mean, if we talk about like millions and everything, sure, but that's like the 1% of the 1%. For the rest of the population like us, I think that just like the joy in the dopamine is coming from what you ship. So like having this ability to ship more value and have more customers, like being happy with what you do. Like you end your day and you feel like, damn, that was a good day. So I think that the discussion is not about like the money itself. but like, oh, damn, I'm in an environment where I ship fast. I can have, like, all the tools that I request within 24 hours.

I can basically be, like, the best version of myself. And I have fun in a good team. You don't have a lot of attrition when I would say you have an environment like this. So, like, sure, money. So you need to pay people like fair amount. But if you're, like, just, like, fair, people tend to stay if you have the right environment. And, like, helping them to go from four per week to six, they're like, shoot. Like, I'm so much better than.

beginning of the year. That's so cool. And you don't have that everywhere. Yeah, I'm with you. I'm curious to see more of the scores about. Awesome. Any parting thoughts? Just generally, your take on AI on the software industry, you've been in this for two decades. Do you think that people should still learn to code? Do you think the junior developer is screwed? Any of those opinions that are common topics? Yes, of course. Of course. You need to learn to code. Like I see this about like a kind of like the switch from like assembly to see. Yeah, it's a higher level.

It's just another level of abstraction. But at the end of the day, you still need to understand how a computer is working. You need to understand how memory is working. Like swaps and all of these things happening on the on the server. Like how a server is working like serverless between quotes. It's always a server of someone. You need to understand the fundamentals to be good with AI. I do believe that AI will do.

only one thing, it will separate faster the good engineers from the bad engineers. If you're good engineer and you're using AI well, you will be an amazing engineer. If you're a poor, lazy engineer and you don't want to understand things that you're doing, AI will make you even worse because you will have the feeling that you get it, but you're not going behind the magic, behind the curtain, behind things and how they work. So I think AI is a blessing for our job. Awesome. Any final call to actions, hiring people, things you want people to do and trying the product and give you feedback on? Of course, try the product. Of course complain to me. If things are not doing, I would say great and they are not great. Yes, we're hiring. So we hiring product engineers. So people that have a strong appetite for like the user experience, because I do believe that in the world where the technical mode is not that a mode anymore because like startups in two weeks, they can build something.

that is close to what you're building. The difference is like how you think about the user, the flow and all of that. So people that have this appetite for nice interface, beautiful product that people love, this is the type of engineers we want. Good engineers, that's a baseline, of course. But like with this spike into like the user experience, even if you're a back-end engineer, back-end engineer but you care about like the latency because it's having an impact on the end user and all of that, this is the type of engineers we're looking for.

And we don't care where you are. So you can be in Patagonia, as I said. Or you can be up north in Canada. We try to limit things to like America's basically. But yeah, just looking for like bright, gritty people that want to have fun. We're seriously fun. Cool. Thanks for joining us, man.

This was fun. That was cool. Thanks for having me.