Embedded - 264: Do It for the Herd

Episode Date: October 18, 2018

Chris Svec (@christophersvec) returns to chat about recruiting for embedded jobs and to help us answer listener questions. Also, he’s looking for engineers to join him at iRobot. Want to get into embedded and don’t know how? We did a show about that: 211: 4 Weeks, 3 Days. Also, there is an EdX class that is popular and a Coursera course that may be useful. You can meet up with Chris at Hackaday Supercon in Pasadena, CA on Nov 2-4. Fulgurites are cooled lightning.

Welcome to Embedded. I'm Elysia White. Today we have Christopher White, as usual, and Chris Speck is here to chat about stuff and things in college recruiting. Hello, Chris Speck. How are you doing? I am doing great. You too. How are both of you doing? Good. Acceptable. Good. Acceptable. That's as good as it gets right there, isn't it?

You've been on the show a couple of times, a couple of our group shows, but also you were on Happy Cows talking about empathy-driven development. For people who haven't heard that or any of the others, could you introduce yourself? Sure, sure. So I'm Chris Feck. I'm an embedded software engineer and a recovering chip designer. I've been doing this for quite a while, and I enjoy the know i enjoy the job the industry and then the kind of meta job that goes along with uh with talking to you all and listening to your show and uh kind of doing this doing this in our spare time so yeah i'm happy to be here so you're saying our show is a job for you it is a job for me it's a very low paying job but it's definitely a job I pay him in stickers. Okay, lightning round.

I'm not even going to explain because you know. So, Chris, you start. Can you give us a fact about lightning? It's electricity. Can you give us an interesting fact about lightning? There's this thing called ball lightning that I attempted to read a sci-fi book about that was maybe partially factual, and then I got bored and put it down. All right. I don't know much about lightning.

Favorite editor? Vim. What's a VI command that everyone should know escape colon q or escape colon q exclamation point if you know one colon q escape colon q you can't just walk around that's that's true that's true escape colon q that's that's sadly that's true if you were a character on the shows we're about to list which character would you be on the show okay okay star wars luke the good place chidi for us to development place. Chidi. Arrested Development. Michael. Strangest animal you've ever touched.

I'm trying to think. I'm just thinking... Is your brain not organized to like, strangest animal you've ever touched should be like... No, it is not. It's not order one. This is definitely... We're N squared here. This has got a really bad worst case and this question is the worst case um i don't know what if i uh you know the normal stuff but like well then it's not strange that's true snakes i've probably touched a snake before i i'm totally coming up blank oh i i have swam swam i have been in the ocean with

stingrays and got to have them swim up and touch me, and I got to pet them, and that was crazy. Okay. Yeah, those are cool. All right. We'll give you that. That's better than your lightning fact for sure. That definitely took me a while to page that one back in.

Given a free afternoon in Pasadena, would you spend it at JPL or Huntington Gardens? I don't know what Huntington Gardens is. How did you not know that? Half of the good places shot there. Is it? Yeah. I've heard the podcast. I should know that.

I'd probably go JPL. Yeah, because you don't know what Huntington Gardens is. You don't even know what you're missing. That's true. We've only met via computer. Are you real? Yes. As far as you know. I't are you me i am not you but we have

a lot of similarities not just our first names all right worst fictional robot worst fictional robot um goodness you're gonna have to cut a lot of silence out of this one chris sorry man no extending it yeah it took him 15 minutes to answer that it could it definitely could i'm totally like don't you usually give the guests the lightning questions ahead of time oh you don't no it's questions may include but are not limited to ah Ah, okay. What's a bad fictional role? You know what I'm going to say. I have never cared for C-3PO. I'm sorry. I think he's a pain. One of my favorite scenes was when C-3PO's head came off

and R2-D2 dragged him through the sands. So I'm there with you. I'm just realizing now that actually, like, Chidi from The Good Place is actually very similar to c-3po like anxious like overthinks everything speaks more than one language that you know maybe maybe this has told me something about myself probably really does have a garbage disposal in his stomach and a fork yeah it's it's always hurting him archie didn't drag him across the sand he pulled him out of the sand but he was intact his head came off in cloud city wasn't there some net there's a prequel a prequel well it's a prequel

okay okay uh i'm gonna stop lighting around now um if you haven't seen the good place you go see it. You should not learn anything about it. You should just watch it. Yeah, don't Google it at all. Seasons 1 and 2 are on Netflix right now. Fantastic show. And I think all three of us are having a really good time, not only watching it, but talking about it.

Yeah, this podcast is going to be about TV today. Actually, Svek had a small statement regarding his lack of ability to speak for iRobot which i thought was sad because i thought he was iRobot no i'm pretty sure i'm pretty sure he is a robot oh right yes he said i am a robot but i do not speak for iRobot even though I work for iRobot. I just feel like since we're going to talk about some hiring stuff and career fairs, I should just make that explicitly clear so that if anyone from my company hears this, I'm doing my legal duty that I don't speak for iRobot.

I'm just some random guy who happens to work there and enjoy working there. But I could be lying as far as any of you know, and it does not represent iRobot at all. So what iRobot secrets can you tell us? Um, we make robots? Doesn't seem like a secret. Yeah, that's as good as lightning facts, man. Yeah.

Do you make ball robots? At least this one was definitely a fact. Do we make ball robots? No, you're thinking of Sphero. I know, I love Sphero. Okay, so what is it like working at iRobot? do we make ball robots? No, you're thinking of Sphero. I know. I love Sphero. Okay. So what is it like working at iRobot? It is a pretty good gig.

I actually just passed my five-year anniversary there a couple weeks ago. So I have been there five years and it's been interesting work the whole time. We, yeah, we make robots, you know, Roomba is kind of the one that most people know of. That's kind of our bread and butter robot.

We have, wait, you make robots that butter bread? We do. We've been using the Roomba wrong. You definitely have. Well, see, if it finds butter or toast on the ground, it will apply one to the other 50-50 chance. So it's pretty revolutionary. We've got a lot of IP in that area. But yeah, so we, you know, we are what I would

consider a kind of a pretty traditionally set up medium sized company, you know, we're like 1000 people, we have about 400 engineers, you know, 100 plus engineers are software engineers. And then we have, that's our biggest single discipline. But then we have a bunch of electrical engineers and mechanical engineers and systems engineers. And, you know, the four disciplines there, plus user experience and kind of industrial design. We all work together with a product group to try to make new and interesting robots that work really well across your house and across everyone's houses, across the entire world. Which is, as it turns out, actually the most interesting part of the job, is that making a robot that just works in a lab is pretty good. I mean, it's definitely difficult, but making a robot that works in every single house

with every single random floor transition from floor to carpet or door jam that is two inches tall for some reason in New England and tatami mats in Japan, which is this kind of paper-like floor substance. Not floor substance., not floor substance, that sounds like it's a chemical you pour out of a bottle. No, but they're floors that are kind of like a paperish, a thicker paper. And so that's very different than a hardwood floor or carpet. And so just dealing with all these different real world things is frustrating at times, and it's fascinating. And just the number of interesting problems that we get to solve,

and that we get to come up with. Sometimes conventional solutions to, sometimes wacky solutions to is a good time. I've been pretty much fully interested in my entire five years there so far. So, yeah, it's a good gig. You should come work for us. We're hiring. And you're hiring a lot right now, aren't you? We're hiring a lot.

Tons of software engineers go to our website, iRobot.com, and you can see we're hiring software engineers mostly in our Bedford, Massachusetts office, which is just outside of Boston. And we are also hiring in our Pasadena, California office. And you can feel free to email me. Even if you don't think you're super qualified, email me and I'm happy to talk to you. Not just Chris and Alicia, like anyone who's listening to it. You too. You too can feel free to email. I have a job. You've also been going to career fairs. Yep, it is that time of year. Schools, I mean, colleges are starting up again. And so I spent two weeks basically doing full-time recruiting stuff for work. You know, I'm normally an embedded software engineer, but I also actually really enjoy recruiting. I really enjoy talking to students and talking to candidates.

And so a bunch of us went to a few different colleges. I ended up at Carnegie Mellon most recently. And yeah, so we might talk a little more about that today. Yeah, I do. I want to talk about it because I know you've been doing it. You've been in the field, and it's good to get a perspective from an engineer about how it works. But we're not going to talk especially about recruiting and hiring

because next week we're also going to be talking about recruiting and hiring because it's just that time of year. I mean, college students need to know, and everybody else is looking at their Christmas bonuses going, this place sucks. I want to work somewhere else. So, yeah, there'll be more about jobs soon. But first, how does iRobot find students and how do they decide what colleges to go to?

And you've been to some other ones. And yeah, how does it all work? It is, it's complicated. I mean, I won't even speak just for iRobot. I'll try to speak more generally. I've worked at other companies, big and small. And, you know, most companies tend to have a number of colleges they like, whether because their founders went there or, you know, half of the software engineers are from there or because, you know, they are the 15 colleges or the 10 colleges that are nearest where the company is headquartered, those tend to be pretty common answers as to why a company goes to recruit somewhere and not somewhere else.

iRobot in particular is based just outside of Boston. And when did we get founded? We were founded in 1990. Three MIT, well, I guess one professor and three and two of his students went out of MIT and founded iRobot. So, you know, the company came out of MIT, so we recruited MIT. There's a bunch of other good universities around the New England area, and so we tend to recruit at WPI, which is a school up there. Northeastern, a variety of other schools up there. We also go, because we are a robotics company, we go to the big robotics colleges, which is basically Carnegie Mellon, MIT, Georgia Tech are the big robotics colleges. We found a lot of good people there.

They have good programs and, you know, they're historically very popular as a robotics place. So, you know, that's the way we do that. We have alums at other schools that have, you know, pushed us to go recruit at their schools as well. And even if you don't go to those colleges, we still accept resumes from anybody through our website. But, you know, we only have a certain number of hours and a certain number of weeks in the year. And so we focus on those schools so you spend most of your working hours writing firmware and software and dealing with system things and testing why why are you doing recruiting why you specifically uh is this a good use of your time is this a good use of your time

is this like that scene from office space are you two both bob here um i have uh i got involved in Is this a good use of your time? and I have been fortunate that I've enjoyed the jobs that I have. So I like, you know, kind of selling the company that I'm at. That's a good kind of a selling that's like telling people why I enjoy it. And, you know, the different sorts of options that a company might afford and why you should work here or, you know, why we might not be a good fit for you even, which sometimes happens too. So I just naturally enjoy that. Engineers, as a stereotype, typically do not enjoy interviewing and selling. And, you know, I see the value in that in terms of leverage,

you know, you hire one person, and you, you know, you get hundreds and thousands of hours out of that person's work. So in terms of leverage, I see it and in terms of leverage, I see it. And in terms of enjoyment, I enjoy it. And apparently my management thinks I'm good at it. So yeah. So I volunteer to do it and they haven't asked me to stop yet. So something's working. I did it for HP when I was first out of college and would go back to my alum and interview there. And I had so much fun because the college students were just so awesome. It was so fun to see their enthusiasm for everything. Well, in this most recent visit to Carnegie Mellon, so I graduated, I finished my master's degree in 2002, right?

And I went to Purdue, a great college, had a great experience there but it's funny because in 2002 the robotics algorithms that are currently being taught like at an undergrad level or at a you know intermediate grad level weren't that they were phd material in 1999 and maybe they didn't even exist yet and machine learning was something that was you know a silly pipe dream that would never really work in 1999, in 2000, 2001, when I took those classes. And now, you know, at Carnegie Mellon a couple weeks ago, these students are coming up to me who did stuff that was literally not possible. And they're undergrads.

It was literally not possible in 1999, 2000, 2001. And they did it in a semester class, you know, in the fall. And so that is just amazing and pretty cool and reminds me that I am, in fact, getting older, but that it's a pretty cool time to be in college, a pretty cool time to be working. And it's pretty cool to hear what they are learning because it means that that material is out there.

It isn't as hard as it used to be, and you can go search for it and maybe learn it yourself. Exactly, exactly. And, you know search for it and maybe learn it yourself. Exactly. Exactly. And, you know, you get things like Udacity. And, you know, I know that you certainly, we talked about this before. And you've taken these courses that, you know, all this stuff is reasonably accessible. It's reasonably affordable depending on, depending where you're at.

A lot of it's free. And so, yeah, it's just a crazy amount of information out there. So I remember as a student interviewing and not doing that well a few times. What makes for a bad interview from your perspective? Hmm. A bad interview in terms of like a college fair kind of a thing, like the, you know, five minutes or in terms of like a real interview, like, a you know 45 minute to an hour interview interview um i guess we did 45 minute ones uh we didn't do the five minute thing we didn't do so maybe tell us how it's structured oh sure sure well there's there's two phases i'm thinking of so one is the career fair you know where it is a usually a room like a gymnasium at a college and 30 50 100 200 come in. And for one or two or three days, every company gets a booth, you know, a couple of tables, you put your little sign up and back you. And so you

have all of the undergrads and grads get released into this gymnasium or whatever, whatever facility it is at the company or at the school. And then basically it's a, it's a free for all for, you know, 8 to 5 or 9 to 4 or whatever it is. And students line up at each company's table that they want to speak at or speak to. And then, you know, companies have one, two, three, four or more people there. And then you get, you know, a few minutes per student basically, but it's the quick kind of matching thing. So they'll come up and they'll, you know, a student will come up and say, hi, I'm, you know, I'm Bill.

I'm interested in robots. This is the kind of thing I think I want to do. And then I, as the recruiter, will say, okay, well, we have openings for three, four, five different types of software engineers. Try to do a quick, you know, one minute overview and then say, you know, do any of this sound interesting? And then you, if you can, you even ask maybe a technical question if you have time.

And if it kind of works, if it doesn't, you just hope to make a good, super fast appraisal. And then you go back to campus or you go back to work the next day with 300 resumes in your hand. And you attempt to remember who were the students that might have been a better fit and try to go from there. But that's the career fair thing. I haven't decided if they're actually inefficient or very efficient, but they are definitely a very unique thing. Sounds like going to a conference for cattle. You know, they're typically held in gyms.

Everyone is kind of sweaty because there's way too many packed in, way too many people packed in there. And, you know, the students are kind of nervous. Maybe the recruiters are kind of nervous. Everyone's standing there the entire day. So it is a physical, like, it's an event. It's not an Olympic event anytime soon, but it is definitely something you got to have endurance for. And, you know, some of these things, they go on for one, two, three days, and man, students, my hat's off to them because they got to stand in line for half an hour to an hour to talk to some companies. And if the company's, you know, kind of one of the big ones

like Google or Amazon or Microsoft or Apple, you know, the line can literally be an hour or two long. And you're just standing in line, you're waiting and you're waiting and you're waiting and you're hoping to talk to a recruiter and you're hoping to make a connection in like two minutes and then next. That's insane. That doesn't sound efficient.

Take a number, people. It's insane. It's insane. If you're a student and you're looking to stand out, the better way to do it or a good way to do it is to find out some sort of a club or activity that you enjoy. For instance, if you are interested in iRobot, look to see if your college has a robotics club of any sort. And then you can email someone at iRobot, someone like me,

and say, hey, I am part of Carnegie Mellon's robotics club, and we would love to have you come talk if you guys are coming to campus this semester and then that is a really good way for us to first of all have kind of the inbound interest and then you get a couple of company representatives to come to your robotics club or your chess club or your underwater basket weaving club whatever it is and you know that it's a much better much better ratio of people you can actually have conversations you can actually like you know, talk for more than two minutes and you can talk one-on-one or five-on-one or whatever in a much more meaningful way. But we do the career fairs and we get good people from career fairs. And then we do the individual stuff, you know, with the student clubs.

And we also get good people from that. And we get good people, you know, just cold emailing us. And, you know, we have recruiters that do all kinds of work to source candidates as well. Okay, so back to my question, and now I will refine it. I still don't know anything interesting about Lightning. I have not had time to look that up. In a five-minute career fair interview, how do you bomb it? You bomb it by not knowing what the company is that you're talking to,

but not knowing what they do. Oh, yeah. Is this Urbop? Yep. Or you come up to iRobot and you say, what do you do? And you clearly don't know anything about us. Or when the company describes what you do, you snicker derisively. Yes.

This was a mistake that I made in a college interview. Oh, really? What company was that i'm not gonna say because i think they still exist um but what did they make i don't know i that they were it was the job the company wasn't so bad but the job they were hiring for i think this was for summer internship jobs i know which company this is but it was some the job was it seemed very data entry it was data entry kind of scripting thing yeah uh okay anyway but that did not help me yeah don't

don't don't laugh at the recruiter you can laugh with them but don't laugh at them so yeah that's that's a good good thing but yeah not knowing anything about the company you know you don't have to have a phd in iRobot right but you should know the company that you're talking to. What kind of products do they make? What kind of people do they hire? Do they hire software engineers or are they a chemical components company? Are they in a location that you want to move to or intern at or whatever? So you should know some of that stuff.

It's nothing that a five or ten minute Google search wouldn't tell you for any company. And I get it. You know, you're on campus, you're doing the career fair, you're a student, and there's a line that's not very long. So you go and hop in it because you're tired of standing in two hour lines. I totally get it. But use that time to do a quick search on your phone to figure out what company it is you're talking to. So that's one way to just kind of bomb and not, not impress us. The unfortunate thing is we get like 300 resumes in a day, right? And the, it is a horribly, horribly inexact science. Recruiting on both sides of the table is horribly inexact. And so if, if any signal kind of comes up that appears to be positive or negative,

you're going to take it. And if someone doesn't seem to know what I wrote, to care enough about your company to even do a quick Google search for it, then it's kind of like, well, I'll talk to the next person instead. So it sounds a little callous, but that's an important signal you can send to a recruiter is, I know who you are. I know who your company is. I know that you have a job or two that sound interesting. And I think I'm qualified for that job. And let me tell you about why I am qualified for that job. Yeah, because it cuts out four and a half minutes of the recruiter giving the same speech they already gave 97 times.

Oh, man, if you come up to me, if you come up to me at CareerVarian and say, hi, I know that you guys are hiring for, you know, a new college grad. People who know robotics algorithms. And I love the Roomba. And my grandma has a know robotics algorithms, and I love the Roomba, and my grandma has a Roomba, or I love the Roomba, I'd love to use one, but I've never seen one, you know, in person. Can you tell me more about what the job is like, or what work in there is like? Like right there, you just went up to at least the top half of people I've talked to, because you're interested, you know something, and you're clearly interested. So that, you know, within 30 seconds right there, you've stood out a little bit at least. Okay, for the longer interviews, the 45-minute ones, what are some things people do there that you see that makes you feel sorry for the interviewee. Yeah, yeah. So people who, it really, really, really, really helps to be able to think out loud when you are being asked questions.

So, you know, if you're asked a question, whether it's a coding question or anything that's not just a simple sort of back and forth, you know, conversational thing. But you have to go off and think about an algorithm or do something that takes more than just, you know, an immediate recall kind of an answer. It is really hard to sit there in a silent room with an interview candidate. Um, if, especially if they're not like sketching something out or drawing in the whiteboard or, or doing something like that. So if, okay. Cause then you're not sure as the interviewer, you're not sure if they are, if they understood you, if the candidate understood your question, if they're stuck, if they need like a little push, if they, you know, have just like totally shut down, you just don't know what's going on. Or if they're a genius and they're thinking it all out in their head, and then they're going to come up with the perfect answer in two seconds. You have no idea

where people are in that spectrum. So if you can bring yourself to get used to talking through problems, that will help. Perhaps that will help you feel more calm in an interview and it will help the interviewer kind of gauge where you're at. Because usually if you're asked a question, the interviewer, a good interviewer, isn't going to be like, here's a question. I will give you no more data about it. You must come up with an answer and I will not talk to you again. That's usually not how it works. Usually you ask a question, hey, how do you think X might work? And then if people are kind of confused, you say, well, let me tell you a little more about there. Let me clarify. Or if you ask a question as a candidate, then I'll try to explain something

that I may not have explained right the first time. So you love the back and forth just to see how people are doing. I've had interview questions where I don't give you all the information. You have to ask. And if you don't ask, if you just make the assumptions, then that's, you know, that's, oh, well, okay, that's not the way I wish I would have solved it. But yeah. And yeah, you have to be able to talk aloud although you can sometimes say i don't quite know how to solve that problem let me solve this one that i do know how to solve and is similar and you can get some points that way and sometimes that talking through the similar

answer that you are already pretty confident with will lead you back to the problem they asked for i've always found that really helpful sure i don't know that i've ever seen it before but i think i that you are already pretty confident with will lead you back to the problem they asked for. I've always found that really helpful. I don't know that I've ever seen that before, but I think that's a great strategy. Because, yeah, we have a list of questions that we use. And if you take a different tact on an answer that we've never seen before, we're happy to follow you,

partially because that's interesting. It's not like the 73 other people we just interviewed who we asked the exact same question of. And partially because, yeah, we're just trying to see, you know, can you solve these software problems? And if you can solve a similar one and you just, your brain didn't quite figure out how to solve the one that I originally asked

on this given day in this given room at the end of a six hour interview, you know, hall, then that might be fine as well. So yeah, that's a good suggestion. Thanks for being on the show. Actually, we have Patreon Slack questions from the people who hang out in our Slack channel who got there by being Patreon supporters. Are you ready for those?

You've seen them all because you're in the Slack. I'm in the Slack. I don't remember them though. Okay, here's the first one. Can you give us a fact about lightning? That could have very well been in there. I don't know anything about lightning. Ben Franklin invented it when he was, I don't know. Okay. um when he was when he was i don't know okay i was i was i was on i was on this is totally random i was on uh vacation a few months ago and we were in um i'm one of these you know these bus tours around the city and uh the the person was actually talking about ben franklin i forget what it was and they said he invented lightning back in whatever year it was with his kite. And I just laughed, and my wife gave me a dirty look.

I mean, that's a crime. If he invented lightning, it's not a good thing to have around. Huh? He was a polymath, so he did a lot of stuff. It's a bad idea to invent something like that. Okay, let's just clear the record here. Ben Franklin did not invent lightning. Okay, and I'm going to go on to the Patreon question.

Why does every job requirement, even for entry level, expect three to four years of experience? Yes, I saw this question and I actually went and looked at our career site to make sure that we did not do that. So that is a thing that happens. It is not good. Companies should not do that. Generally, when you have a job rec, which is short for job requisition or a job ad, when you create an entry, you say, hey, I want to hire somebody and you put it on your website, you generally think, okay, is this person a new college grad? Are they kind of like zero, one, two, three years of experience? Are they, you know, five plus years experience or 10 years experience? You try to kind of roughly bucket it in and job descriptions, the, the, the, the way the job descriptions get written in a company is your internal recruiter. Who's in charge of actually putting the job rec together

and putting it up on the website. They email the hiring manager, the hiring manager ignores them because the hiring manager is worried about product deadlines, not hiring deadlines. And then the recruiter emails the manager again and then corners them in the hall and says, can you please get this to me? And the hiring manager says, yes, I'll have it to you by the end of the day. And they don't. And then the recruiter needs to update something on the website because his boss says, we need to hire somebody for this now and your salary depends on it so they put something together and they're doing the best they can and they can't read the manager's minds and the managers don't communicate very well with

recruiting because everyone has kind of different goals and so that is how that happens this by the way is why it takes three to six weeks sometimes for you to hear back after, even after you think you've nailed an interview or gotten the job. Yes, yes. It's not because companies are mean. It's not because companies are incompetent. It's just because there is a lot of communication that needs to go on. And the recruiter's full-time job is recruiting, but the manager's full-time job is not recruiting. And so, you know, some companies handle this better than others. Some companies like explicitly say, you know, 20% of

a manager's time is hiring, is recruiting, is, you know, selling the company. And so you might see different results at a company like that. But yeah, so sometimes job recs get fall through the cracks. I mean, I've definitely heard things like in the web world nowadays anyway, that companies are basically saying like our sort of quote-unquote lowest tier or like most junior tier of employees have to have at least three or four years of experience and you know that that's just a decision they make they're just not hiring that means they're probably not hiring new college grads at that point um why why they would say new college grad and three to four years in the same experience, in the same job rec, is probably just a mistake or a typo or a couple of job recs that got smashed together and no one looked at them.

That's a good point. Often job recs are just copy and pasted over and over again. And nobody has read it at the company for years. Well, and there are small companies that don't want people with less than five years of experience. Startups generally aren't hiring college grads unless they're further along and established, right? I mean, a lot of small companies,

they want senior people to come in. So, I mean, if you're looking at startups, you might see kind of a bias toward, yeah, five years plus, please. Yeah, I guess I sort of assume what this person meant that I was asking the Slack was something where they're looking for a new college grad who has five years experience in these web technologies. Yeah, right, right. Which is kind of like, well, I mean, yes, there are people like that, but that's certainly not. Yeah, and I remember seeing ads for Java, 10 years of experience in like 2002.

Right. It's only been around for four years or whatever. So yeah, badly written job ranks are more likely cause. Well, and defining experience is tough. If you say you want three to four years of experience, is that three to four years of experience with programming? Because maybe some of the college students have that. Or three to four years of experience with programming? Because maybe some of the college students have that. Or three to four years of experience with C as a programming language and embedded systems, that's a different level.

And so when you read the job recs, sometimes you just apply for all the jobs. Yep, yep, yep. And I mean, that's absolutely fine advice. That's what I try to tell people, yep. And I mean, that's absolutely fine advice like that. That's what I try to tell people too. When we got married, my wife, she's from Indiana. She's a good Midwestern girl. And I would apply to jobs for which I did not meet every single job requirement. A job maybe has like 10 things that you need to have. And I would apply to one, apply to a job that maybe I had five or six or seven or even three. And she viewed it as kind of dishonest. Now my resume only, my resume was 100% true. So there was no, you know, I wasn't like stretching the truth or anything. But if you matched my resume up to the job recs, I would only meet three of the requirements and not 10 of them,

but I would still apply. Because at the time, I sort of guessed this, and now I know it to be true, that, you know, these job recs just get copy and pasted all over the place. They're not hard and fast, you know, legal requirements kind of a thing. So if you are interested in the job, and if you can look at it and honestly say, yes, I think I could do that job, then there's no harm in applying for it. I mean, it takes your time to apply for it, but worst case, the company ignores you. And there's often stuff that's just kind of accreted on, like you said, they could reuse them. There's stuff in those job recs that they don't even care about. Oh yeah. I mean, I've seen some where it's like, having been in this career for a long time, I know what they're actually after, But then they'll have just, oh, and you need to know SQL and this.

And I'm like, no, you don't. I mean, you need to know it enough to look up the man page and type one command. Because, you know, there's a back-end database that your device hooks to. But it's not like you're going to be going in and maintaining this database or something. Absolutely, absolutely. And accrete is the perfect word for that. The job recs, those webpage descriptions of what a job requires

are very... The way that those bullet points add up is accretive. Is that the right... Am I using that word right? Sure. They accrete. They just kind of grow up like volcanic layers of lava hardening. My goodness. I don't know about lightning.

I don't know about volcanoes. I don't know about volcanoes. It's a good thing I'm in the technology field and not the natural field. Okay. Another question is how have applicants stood out to you in the past? And we talked a little bit about actually knowing what you're applying for. Are there other pieces of advice you can give for that? Yeah. Anything you can do to put yourself in the shoes of the employer, of the recruiter is a win. So, you know, if you have a pithy little elevator pitch, that's great. If, you know, something that I've actually seen at a couple of colleges now where two guys, one at Georgia Tech, one at Carnegie Mound, they wore a shirt that had a little bit of purple in it. And one of them had a bow tie.

One of them had a regular tie that was purple. And their resume, you know, normal black and white resume, but some of the words, like the more important words, were the same shade of purple. And so that was, and I actually asked both of them about it. And I said, oh, is that intentional? And they said, yes. It almost sounds gimmicky. It almost sounds like a little of a trick. It was understated. I'm sure plenty of people didn them about it. And I said, Oh, is that intentional? And they said, Yes. And like that, it almost sounds gimmicky. It almost sounds like a little of a trick. It was understated.

I'm sure plenty of people didn't notice it. But it was it's a little thing that made me stop and think and I remember this person. No, I don't remember their names. But at least that that was a little hook. But you could find their resume. Yeah, I can absolutely find their resume. It was a purple guy. And then I think one of them I said, you know, hey, why don't you email me later? It seemed like a good fit. And I said, like in the email, why don't you email me later? It seemed like a good fit. I said like in the email, say I'm the purple guy. That way it jogged my memory after having talked to 300 people that day. Um, if you are part of, if you've done any kind of a project that is what the company does, that is what the job does, then lead with that.

Hey, I have built a, you know, a Roomba clone at home. Hey, I, you know, you want to go work for Google. I built my own search engine just to try, you know, it doesn't work very well at all, but I did it. Anything that makes it look like I have done the work that you would want me to do. That is, again, a no brainer. Cool. There are a couple more, but we're going to get to them later when we get to email questions before we do that

I have two short things first did you get a flu shot I did get a flu shot a couple days ago Chris and I both got our flu shots last weekend and listeners flu shots are important the importance

is not for you the importance is for the herd. Do it for the herd. This is like a mid-show PSA? It's the commercial PSA. Yeah, that was it. And you should do it soon because you need the immunity by Thanksgiving when all heck

breaks loose and your family comes. It is also amazing nowadays how most grocery stores, or at least many grocery stores, have it for free or for dirt cheap and you can just go and get it at any old CVS or Harris Teeter or whatever Safeway. I'm trying to think

of different regional. Piggly Wiggly. Publix. Wegmans. And you are going to Pasadena in early November. I am. I am going for a work trip, and the timing actually ended up working out really well,

where the Hackaday Supercon is going to be there, so I'm going to be stopping by there as well. Are you going to wear an embedded shirt? I don't think my embedded shirts fit me anymore. I tend to shrink all of my clothing. I thought you were going to say I tend to shrink. I do not tend to shrink. In fact, the clothing may not be shrinking.

I might be doing the opposite. No. But, yeah, I have a t-shirt, but it does not fit. It is not a good look. Trust me. You do not want me wearing that shirt there. I'll probably be wearing an iRobot polo shirt or two. But you'll have embedded stickers.

I will have embedded stickers. Unless you forget. Many embedded stickers to give out. No, I should actually pull them out and put them on my bag to take there. But yes, I will be there. I'm looking forward to it. I'm looking forward to Chris Gamble's talk. I'm looking forward to meeting a bunch of people who I've never met before, but know online.

It should be a good time. Chris Gamow will be able to tell us if Svek actually exists. That's right, that's right. Although maybe he doesn't exist. You've met him, haven't you? No, we've met him. We've met somebody claiming to be him. That's true. Chris's are largely

fungible, so. Yes, exactly. I mean, it's all just Chris. Okay, so now the listener email part of the show. Let's see. Let's take the easy ones first. From Mike. Mike says, I am a student at ASU and Professor Soderman is the best professor I have had at the university. It was awesome to hear her on the podcast. That didn't really need to be read.

I just really liked it. Thank you, Mike. No, you guys can comment if you want, but you want an actual question, don't you? I don't have to comment on that. I did enjoy that episode, and actually I'm going to go through, look into her, what's it called?

Robogrok? Robogrok. I looked at the first lecture on YouTube and it looks great. It looks like a very practical and just enough theory and enough practice to actually do it. It looks like a fantastic course. Your show with her was a good one. So, you know, the XYZ coordinates you can make with your hand and for me, I have to have it in order or I will swap X and Y and then I'm not in right hand coordinates anymore.

And so I've always had this way of doing it. And I've given conference presentations that show the inertial navigation gang sign. Her way was so much better and so much more comfortable. I was kind of embarrassed. Do you know where she got it? Like, is this just another school of, of,

of thought that just is like from whatever her background is, that's kind of the way they taught it. And she absorbed it from there and you got it from West coast, East coast kind of a thing, or what is it? I, you know, it was in my list of questions to ask her,

but I ran out of time. Wow. You need a, you need a better producer. Yeah, exactly. You know, it was in my list of questions to ask her, but I ran out of time. You need a better producer. Yeah, exactly. Wait a minute, that's my job. I thought that was Chris. Chris does all of the audio producing, and I do the content production. I got it. Neither one of us does the time production. Ah, I got it.

Neither one of us does the time production because, eh. The time production? We're trying to figure out if we're actually going to hit a certain amount of time. Or if we're just going to go over and talk for an hour and a half. Kind of like this show. Okay, so the next question from Adam, who just started listening to the show and went to college for CS and has been working as a network admin. Has been playing with Arduino's basic stamps and Raspberry Pis. Thought web development might be a better choice to break into the tech industry.

And just wanted to know, how do you get into embedded? And there were some emails from the Patreon Slack for you about what skills should I be focusing on in order to get a job in embedded? I graduated with my degree in electrical engineering. How do I get started looking for a job in embedded? Yeah. So I guess there is some question. How do you get a job in embedded. Yeah. So I guess there is some question. How do you get a job in embedded systems? Good question.

That's the way you stall, right? Yeah, that's exactly right. That is the media training. That's a good question. How do you get a job in embedded? You know, there are many ways. And so far you've said nothing. You just keep talking and talking.

Exactly. So I would say flip it around and look at what the company wants. You know, look at, for any job, look at job ads for embedded software engineer or embedded engineer. You know, go on Indeed or one of these, Indeed.com or one of the big search engines and pick a big city and maybe the city you're in and look at half a dozen or a dozen companies and just go look at their job ads and see what do they want for a zero to five year or zero to 10 year embedded software engineer or web engineer, whatever, whatever field you're trying to get into. And just look at, look at the bullet points. And I'm guessing, you know, you go across 10 jobs and you're going to see probably five or 10 bullet

points that are common across those, you know, for embedded it's C programming. It's a little bit of C plus plus, you know, I definitely see more C plus plus nowadays. Um, but you have to have a solid grasp on C programming. You have to have a solid grasp on embedded processors of some sort. Arduino's are great. Raspberry Pis are great. The work that most companies do, at least in what I think of as embedded jobs, is kind of one level lower than the Arduino IDE. And so you're programming an STM32 chip using C and using some IDE. You're programming some Nordic Bluetooth chip using C and using their IDE or TI chip or any of these chips. So it's a step lower level and more detailed and more detail-oriented and less abstracted, I guess,

than Arduino. So you're looking for embedded microcontroller experience. We're looking for C programming. Python skills tend to be the second programming language used at the companies I've worked at anyway, like the C C++ is used for all the stuff that runs on the device or on the robot or on the chip. But Python is the glue that does all the testing and it does all the, you know, any kind of data generation or any kind of data processing or any kind of debug stuff off the chip. So those, those are kinds of the three, the three skills. And then again, if you can have some project or some experience that says, look, I have done the thing that your

company does that this job opening does, here's this project that I made, or, you know, whether it's a school project or a project on your own, that is a way to kind of stand out in terms of the skills and the experience you have. Um, I mean, you, you both have worked at a number of companies, so you, you've got to have, have some other skills and other,

other things to add to that as well. I hope. Uh, well, I was going to say kind of specifically to, to him, uh, to Adam's question.

Um, he's already a software tester working with embedded devices. So there is a path from there to development, depending on the company. And you can say to your manager or whoever supervises you, you know, I'd like to get into development. Is that something that is a possibility for me here? I already know our system from the testing side. Perhaps starting to pick up some small development tasks.

You know, for small companies, sometimes that's great because it's like, oh, good, that'll take off some load off the developers while you learn. So that's kind of a, you've got a special path there that other people may not have.

The other thing I would say, because you mentioned projects and he's interested in audio, there's a lot of really cool audio stuff you can do with Teensies, starting with Arduino. There's a whole library available, and you can make a little thing that does digital signal processing and have it as your little demo,

and then you can kind of move lower from there. But having something like that, if you want to do something in a particular field, having proof that you've done something is pretty important. Other skills? Well, my advice actually was a previous podcast, the one with Dennis Jackson, Four Weeks, Three Days, number 211, in which we talk all about if you're a software engineer, how do you get to embedded?

What that career change looks like. And talking some on the Patreon Slack channel, not to keep mentioning that, but there were multiple people who voted for an edX class called Embedded Systems Shape the World, Microcontrollers Input and Output. And another vote for the UC Irvine IoT class on Coursera. I will put those in the show notes. But yeah, there are classes out there and the classes

will help you build projects and the projects will help you in interviews. It's not too late it's not too early embedded is interesting well so yeah it hasn't changed that much that's the nice it's a little slower than the web than other yeah other areas so you're not you're not starting out from like a position of disadvantage of having remembered something from 20 years ago and having it immediately completely out of date. I mean, I'm a poster child for that. So I designed chips for the first eight years of my career. And I came out of grad school thinking that's exactly what I wanted to do. And then I did it and it was fine. I got a little bit burned out

and I wanted to try something else. Then I did some projects. I had a friend at a company. I used their, that company's dev board to fool around and to do a little bit of work in the side. And then, you know, I knew my friend at the company and he was able to help me get an interview, which is another good way to do things. But, you know, once you are five to 10 years into a software career, a reasonable recruiting or hiring manager will realize, okay, this person can write software and they have demonstrated that they understand what embedded software is and what registers are and what memory is and how, how one does those things. And so, you know, if you get the right break, someone might take a chance on you. Okay. Next question. Actually, I, sorry, I want, I want to go and do two more

things kind of along those lines. You know, if you are in a company, so, so Chris, um, you know, mentioned that this person was already in, you know embedded test role when you go to your manager with something like that you know hiring is really hard right now you know the job market is really good for candidates right now and so i think you you have the uh perhaps this is a good time to go into your manager and say not just say hey i'd like to do this kind of thing. Do you think it'd be possible, but go in and say, I want to do this thing. How can we make this happen? Like, don't, you know, don't, don't be arrogant about it, but just kind of say, look, this is something I really want to do.

I love the testing. I want to do this. I want to do more. And then, you know, that, that kind of, that kind of is more proactive on your side and you're not just saying, eh, if I could do it, it'd be okay. Instead, you're saying, I want to do this. And obviously you have to have the political capital to be able to say that. And you have to have the confidence to say that. And I realize it's not easy. And maybe you don't feel like you can do that at your company, but if you can, if you can even consider making a strong statement about like, look, this is what I want to do, then, you know, psychologically that will help your manager perhaps be more likely to set you on that path.

The other thing is that if you are very far from the embedded field and you've been a web programmer for 10 years or you're an electrical engineer, well, put the electrical engineer on the side. Let's say you've been a web programmer for 10 years. Try to find a company that does web and embedded. Go get a job as a web programmer there. Hiring someone internally is way easier. And you're doing kind of transfers is much easier in most companies than it is to hire a stranger from outside the company. And so if you can do that, then you get

your foot in the door, then you wander over to where the embedded people sit, and you go have lunch with them. And you, you know, you go and do that a little bit, you get you get known, and they find out you're actually a reasonable human being. And then, you know, you go and do that a little bit and you get known and they find out you're actually a reasonable human being. And then, you know, they have a job opening and they know that you do some of the stuff in your spare time and, you know, good things happen. Guaranteed. Speck's patented method for getting a job in Embedded. That's right. That's right. Okay, so moving on?

Moving on. Austin asks, he's done a lot of embedded development in the past for his master's thesis, where he wants to start doing more embedded projects, but cost is a concern. He's struggling to find a platform, non-Arduino, in which the tool chain is not prohibitively expensive. Do we have any suggestions on devices, programmers, tool chains that would run, say, sub-$100 categories? Thousands. I know. I was surprised by the question. I was like, yeah. So the tool chain, the GDB, GCC, GNU tools chain is $0,

and that works with nearly everything. You can run it under Eclipse. you can run it under VS Code, you can have a fun ID experience. Tool chain should not be a consideration for cost. Even IAR, if you want to do small projects, I think you can get it if you don't care about programming more than 32k of code,

something like that. But I would recommend just going the GNU tools route because a lot of companies are just giving up on it. Most of the companies I've been working with are doing GDB now. And Nordic has a GDB tutorial installation. Yeah. So as far as boards...

Yeah, so then you need a programmer and a board. Or a programmer that's on a board. Yeah, a lot of the dev boards now come with STLinks or JLinks built in that are essentially programmers. So look for one of those. And both of the suggestions from your blog series and Andrei's are, I mean, they're fairly inexpensive, right? Yeah, yeah.

I mean, mine is the MSP430, and those dev platforms are between $5 and $15. And Andrei's is the STM32. I think he's got like an F4, which is one of their more capable boards, and those are $10 or $20. I mean, not knowing anything about exactly what this person wants i would say get an stm32 dev board so one of their discovery dev boards and they come

with the full build environment for if you want to use kyle or ir and pay for that or if it's smaller than 32k flash i think you're right chris then that's free um i believe st also comes nowadays with the what they bought they basically bought a eclipse type of company oh yeah and so they have a gcc um out of the box support as well for for their discovery boards or at least for their some of their eval boards and so i think that gets you that gets you done easily under 50 bucks, maybe 25 bucks, depending on what the discovery boards I haven't bought a discovery board in a few years, I have literally a box of them sitting next to me for a few years ago. But yeah, that would be my first choice unless, you know, if there's some special programming requirements, or you know, if this

person is an analog circuit designer, maybe they want to do more analog-y stuff. And then I don't know that the ST has a ton of sort of onboard analog stuff to play with. But that's my ignorant recommendation. Well, the Discovery and the Nucleo boards from ST are all embed capable. So if you don't want to walk through your code, then you have an online compiler that's really easy to use and has a lot of tools out there for you, a lot of code that you don't have to write. And the MSP430, that uses the Code Composer, which is free now.

Yep. Now, with Embed, I've never used it. Can you actually get, like, oh, I need an I2C library or I need a SPI library in an example? Do they actually have that kind of canned on the web? They've got networking, they've got TCP IP, they have

an RTOS. It looks a lot like Arduino sometimes. Yeah, it's got a ton of, and there's like official stuff, and then there's also community libraries, which are pretty extensive. But you can't step through your code, which, I mean, that's sort of

that's getting away from the Arduino thing is where you can start stepping through your code, which with GCC and GDB you can, and it doesn't have to be command line. There are many much cheaper debuggers out there. Than the JLinks.

Than the JLinks. Blackmagic Probe? Yes. Is's what it's called probe black magic something black magic probe and philip frieden's chip has a programmer for it those sometimes use open ocd which is a kind of a thing that i recommend staying away from because i've been burned by it too many times. Some people have a great time with it. It works just fine. Works with GDB. All's good.

For me, when my tools crash, it infuriates me because I only want the software to crash because I'm stupid, not because my tools are stupid. But that's all getting better. And I recently was given a $4 debugger that was along the same lines of Blackmagic Pro. Heck, if you don't have a computer, you can buy a $25 Raspberry Pi that you can probably use to cross-compile to your target port on and run all of this stuff. You still need a display.

Yeah. Yeah, but you could. This is all, yeah, sub a hundred dollars you have so many options you have money left over yeah money left over for censors or for patreon support of this podcast no don't if you're struggling to buy to buy uh development stuff don't send us any money no that's definitely definitely true um i did have a little bit of back and forth with Austin, and home automation was one of the things that was mentioned. And I think that, yes, under $100, sure.

What was the thing Gabrielle was playing with? That was Raspberry Pi based, right? It was Raspberry Pi was the server for ESP32 and ESP8266. That's another set of things you could get into. And those were doing MQTT.

That's it, MQTT. Which is a very popular home automation system. And those ESP32s are $4 if you buy them online in places and $4 if you buy them online in places, and like $10 if you buy them from Adafruit and Sparkfront. Those ESP32s have Cortex in them, so if you want to get into Embedded, great option.

I don't know how to debug them. I don't know anything about them. Their serial interface is the debugger, but I've never played that much with it. Okay. I think that's actually maybe it for questions. Okay. Chris, do you have any questions? Do you have any questions?

Many about lightning. How does lightning work? I'll just go look that up after the show just to have something interesting. It gets to be a charge imbalance between the surface of the Earth and the cloud. And eventually, it reaches the breakdown voltage of air. Which is many, many thousands of volts. I think the breakdown voltage of air is like 400. But most lightning is much higher.

And I think if I had to give a fact about lightning, it would be about Fulgurites. About what? Fulgurites? You two are the same. When lightning strikes sand, it heats it and makes glass. And it makes a glass imprint of the lightning bolt. So it's not just a pool of glass. It's like a tree of glass.

I have to interject. I'm completely wrong. 400 volts is obviously not the breakdown voltage of air. That is way too low. It is like 15 to 30 kilovolts per centimeter. Thank you. Sorry. I don't know why I had that in my head.

Glass. It's really pretty. And I gather in North and South Carolina beaches. You can just find them sometimes. You probably could out there. I've never looked for them. I try to stay away from lightning storms.

For somebody who doesn't know anything about it, you seem really afraid of it. I should die. That doesn't make any sense. That's great. That's awesome. Yeah. Do you want more questions?

Do you want more lightning facts? Because I could go on. I'm hungry. All right. Let's close up with one more question for Svek. What did you want to be when you grew up, when you were young? What was your intended career path after you finished wanting to be a fireman slash ditch digger slash whatever five-year-old truck driver you wanted to be? Professional musician. So I played clarinet for quite a while, which is, of course, the coolest instrument in the world. And then I picked up bass guitar.

And somewhere in the middle there, I thought, I'm going to be a professional musician. I'm going to go to college. I'm going to study music. I'm going to get a job playing in symphonies and playing in jazz and rock bands. And yeah, long story short, ended up in engineering. That's a bit of compression, but okay. Yep, yep.

I also had a Commodore 64 from the time I was eight years old, and I was a prototypical geek child as well. So basically no sports in my household. It was all either music or computers. So I had to end up in one of those fields. I think you chose the more lucrative. Definitely, definitely. Also the one more likely to have a reasonably stable life.

So yeah, I was very close to moving to Nashville and actually doing studio music after I graduated college. And I lived in a van touring with a band for a couple of summers. And that, that cured me of being a full-time musician pretty much, pretty much right there. Yeah, I could see that. All right.

Do you have any thoughts you'd like to leave us with? As you are interviewing, whether you are a new college graduate or a very experienced pro, I think the best advice I have is just put yourself in the shoes of the person who's interviewing you, have a little empathy for them and figure out what it is they're looking for. They want to hire you on the spot. And if you can say exactly what they want to hear,

then you will continue getting the next round of interviews and you will eventually get hired. Do not literally put yourself in their shoes. Do not. Do not like that. If you're interviewing for Nike, maybe that actually is the right move. But if you can figure out what kind of people they want, what kind of skills they want, and then if you have those skills and if you can show that you have those skills, that

is the best advice. I guess the other piece of advice I have is don't be afraid to cold email people through LinkedIn or through whatever in a polite and interested way and non-stalkery way that says, Hey, I noticed that you work at iRobot. I am an embedded software engineer, or I am, you know, I enjoy robotics and I'm wondering if, you know, I saw this job opening. I wonder if I could talk to you more about that. Like I will answer cold emails from strangers on the internet if they are nice and if they are professional and if they are, you know, ask me for a half an hour of my time and a phone call, um, happened about a year and a half ago.

And we actually ended up hiring the guy because he sent me a nice email through LinkedIn. And he said, here's some times that I can talk for half an hour. Um, you know, thanks for reading this email and worst case, I would have ignored it instead. I said, sure, let's talk. And we ended up hiring him. And actually then he brought over his, his former colleague as well. So it's, um, there's, as long as you do it in a nice and professional way, um, generally people enjoy talking about their jobs and are happy to attempt to recruit more people who also might enjoy their jobs. So that's two tips. And when you're hiring, it's very exciting if people throw themselves at you because hiring is very hard.

Yes, it is. Yes, it is. And you go to the top of the pile if you have inbound interest, as they call it. Okay. Where should they send their resume and letters for you to peruse? Well, please email me at my work account, which is csvec at iRobot.com. And I imagine this will be in the show notes as well, but it's csvec at iRobot.com. It won't be in the show notes.

Email in the show notes usually gets too bad. Oh, yeah, yeah. On the other hand, if you don't remember Chris Speck's email, which is his first initial and his last name, at iRobot.com, you can always hit that contact button on embedded.fm and tell me you just want me to forward this to Speck and I will do so.

Cool. Thank you. Our guest has been Chris Speck, ruler of the Roombas. Thank you for being with us. Thanks for having me. This was a good time. Thank you to Christopher for producing and co-hosting and thank you for listening. You can contact us at show at embedded.fm or hit that contact link. I already told you about now a quote to leave you with

from James Baldwin. You think your pain and your heartbreak are unprecedented in the history of the world, but then you read. It was books that taught me that the things that tormented me most were the very things that connected me with all the people who were alive, who had ever been alive. Embedded is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California. If there are advertisements in the show, we did not put them there and do not receive money from them. At this time, our sponsors are Logical Elegance and listeners like you.