Daybreak - How facial-recognition firms turn public deals into private gold

Episode Date: February 10, 2025

Indian startups are making the most of the increasing demand for surveillance by securing high visibility government contracts. But while these can boost a startup's profile, government projects are unpredictable and often difficult for smaller startups to win. As a result, there is a shift underway — private clients are becoming increasingly crucial for profitability. This divide between public and private contracts is forcing these surveillance startups to do a fair bit of monkey balancing. How are they pulling it off?Tune in. **This episode was previously published in December, 2024Daybreak is produced from the newsroom of The Ken, India’s first subscriber-only business news platform. Subscribe for more exclusive, deeply-reported, and analytical business stories.

Hi, this is Rohan Dharma Kumar. If you've heard any of the Ken's podcast, you've probably heard me. My interruptions, my analogies and my contrarian takes on most topics. And you might rightly be wondering why am I interrupting this episode too. It's for a special announcement. For the last few months, I and Sita Raman Ganeshan, my colleague and the Ken's deputy editor have been working on an ambitious new podcast. It's called Intermission.

We want to tell the Sita Ramancahans, my colleague. secret-source stories of India's greatest companies. Stories of how they were born, how they fought to survive, how they build their organizations and culture, how they manage to innovate and thrive over decades, and most importantly, how they're poised today. To do that, Sita and I have been reading books, poring over reports, going through financial statements, digging up archives, and talking to dozens of people. And if that wasn't enough, we also decided to throw in video into the mix. Yes, you heard that right. Intermission has also had to find its footing in the world of

multi-camera shoots in professional studios, laborious editing and extensive post-production. Sita and I are still reeling from the intensity of our first studio recording. Intermission launches on March 23rd. To get an alert as soon as we release our first episode, please follow Intermission on Spotify and Apple podcast or subscribe to the Ken's YouTube channel. You can find all of the links at the ken.com slash I am. With that, back to your episode. Okay, so here is a funny story. In 2021, a startup launched a pilot project for surveillance within a temple premises.

The project was meant to use facial recognition technology to identify token holders at the base of the temple hill and then matched their faces when they reached the top of the hill. Fairly simple, you would think, right? That is what I thought too, but I was wrong. You know why? Take a while guess. You see, temple goers or devotees traditionally shave their heads before reaching the particular temple top.

And that makes their faces look different. So it is very difficult to identify them. The project, though, was not really a failure, but it is on pause as of now. The startup's founder told my colleague Abiramjiji that they are still waiting for government funding or some support from a donor. Why am I telling you this story today?

Well, it is to give you a sense of how we as a country have become a bit obsessed with surveillance tech, including the government, of course. To give you a sense of the scale that we're talking about, earlier in 2024, a subsidiary of the Indian Railway that designs and deploys information systems for the railways floated a tender. This tender called for the installation of nearly 3.5-lac facial recognition cameras

in more than 40,000 railway coaches. What was the purpose of this? We spoke to Kay Sriunivasin, the founder of Algo Vision, a Bangalore-based startup that works in the security and surveillance space. He told us that usually these tenders are floated after some incident takes place and government officials take notice. For example, in October, bars across Maharashtra were mandated by the state government to install facial recognition cameras after a horrible incident of underage drunk driving.

On paper, it all sounds elaborate. Facial recognition, AI-powered traffic management, but in practice, the results are often glitchy. And in the case of the Indian Railways' mega project, Srinivasin told us that there was no clarity either. about the purpose and the goals of placing these insane number of cameras in railway coaches. I'm not even going to get into how dystopian all of it is. We'll save it for another day. But here's the thing. Government contracts are big and flashy,

the kind of deals that, you know, look good on a startup's pitch deck. But they are also a logistical nightmare. This is exactly where the private sector steps in. smaller deals maybe, but with actual money on the table. But this divide between public and private contracts is forcing India's surveillance startups to do a fair bit of monkey balancing. Welcome to Daybreak, a business podcast from the Ken. I'm your host, Nickda Sharma, and I don't chase the news cycle.

Instead, every day of the week, my colleague Rahal Philipos and I will come to you with one business story that is worth understanding and worth your time. Vivek Manoharan, the founder of Surveillance Startup Bipolar Factory, told the Kien that government sector projects talk about accountability and transparency, but they are complicated to execute. They pay peanuts, take forever to roll out and frequently stall halfway through, which is why the attraction to private players. Like the founder of VisionBot, a Bangalore-based video analytics startup told us that technically all their revenue comes from the private sector.

In a private sector project, a single camera may cost about 40,000 rupees. But according to Sri Nivasan, a similar project for the public sector may involve cameras that cost only 10,000. These costs include not just the price of the camera, but also the software built in by the provider. Because, and we all know this, the most important thing for government is to keep the costs low. So it will just appoint the bidder that quotes the lowest price. Which explains why most of the public sector projects don't even move beyond the pilot stage, like the temple story I told you earlier.

In fact, many of these projects are either done pro bono or for a token amount. A surveillance startup founder told Abirami, my colleague, that they've done quite a few pilots for various cities and state departments. and only one or two of these were paid and the amount was nominal. And most often government contracts are taken by industry giants like Infesus, Gio and Larson and Tobro or LNT. These big companies act as system integrators. VisionBot's founder told us explaining how it is only from here that reverse bidding is done.

That is basically when smaller vendors are brought in for specific roles like campus, production, tech management, cable laying, implementation and so on. And in this process, the startup that wins the bid ends up losing margins. Plus, government clients aren't exactly easy to work with either. For example, the founder of an AI surveillance company told us how a smart cities project that their company worked with was not happy with their work because there was an internet connectivity issue at their end. But they were the ones who were asked to improve.

prove the speed of their AI so this issue could be fixed. And with the lowest bidder winning the deal, it is harder to do a fair job of implementing a camera system that works well and meets standards. Now, typically, software costs less than 10% of the total contract value, since once it's built, it can be replicated for multiple clients at a low cost. And this is where they make their largest margins. cameras form about 30% of the cost, with the rest going into implementation. It is easier to cut corners here and provide substandard service in the interest of saving money,

since camera and implementation costs are highly variable depending on the quality of the work. Next comes the critical issue of maintenance. Vivek Manoharan, the founder of the surveillance startup Bipolar Factory, explained it to us. He said many of the cameras that governments enthusiastically install end up getting neglected. Because the software of these cameras needs to be trained on new data sets every few months to keep them up to date. They include a 10% maintenance charge in the budget for this. But government officials do not see the point of this, since they see maintenance as a thing done to physical components and not to the software. In the next segment, I'll tell you why these startups prefer private players.

Private sector clients are much more understanding with the kind of constraints that limit surveillance projects, like the nature of the technology being used and what can be realistically achieved. They are also willing to pay more where it counts. Camera quality, for example, is a big consideration for private companies. It is like the difference between paying for an iPhone and paying for a budget phone. Srinivasan told us that the government would never pay for that quality difference. Most importantly, private sector companies have a clearer vision of what they want to achieve.

But despite this perception that government tenders are tough to bid for and the process is long and complicated, the private sector is no less. Private sector contracts require these startups to put in all the legwork. This involves generating leads, meeting with clients and using connections. Plus, in government contracts, big companies or system integrators like LNT and Infesus do the heavy lifting, navigating numerous hurdles and convincing government officials to adopt realistic, effective solutions. Private contracts, on the other hand, requires surveillance companies to take on all the work themselves. But private projects are usually smaller. VisionBot's founder told the Ken that in his experience, contracts floated by private companies are purely

need-based and are not fully transparent in their aims and scope. Plus, if management changes, the project is likely to be put in cold storage. On the flip side, though, large government tenders have such scale that even offering tech at a steep discount results in a balanced profit by the end of it. In the private sector, though, most projects are between the range of 1 to 10 crore rupees as compared to government projects that usually run in. into multiple crores. And even these are carried out in phases.

Yet, with all the issues in public and private deals, public surveillance is here to stay. Daybreak is produced from the newsroom of the Ken, India's first subscriber-focused business news platform. What you're listening to is just a small sample of our subscriber-only offerings. A full subscription unlocks daily long-form feature stories, newsletters and podcast extras. To subscribe, head to the ken.com and click on the red subscribe button on top of the Ken website.

Today's episode was hosted by Snigda Sharma and edited by Rajiv Siyah.