[Podcast] Brains behind AI's Ari Yacobi discusses Contactless Authentication with Mohit Garg, CEO Oloid Inc.

Mohit shares his views on contactless authentication and how Oloid is using AI to enable the new normal post covid-19



Following is the transcript from the interview of Mohit Garg, CEO Oloid by Brains Behind AI's Ari Yacobi

Ari: Hey, everyone! Welcome to Brains Behind AI, a show where we meet the innovators, entrepreneurs, and the real brains behind some of the most successful AI startups. We ask them about their journey from coming up with the idea to finding product market fit and from their experience draw a set of principles that we can take away from ours. This is your host, Ari. Thank you for spending time with us. And now let the show begin.

Hello and welcome to another episode of Brains Behind AI. I am Ari Yacobi and I have with me my co-host, Natalie Thomas. Natalie, how are you?

Natalie: I'm doing well Ari. Thank you for having me. Today, we have a special guest with us. His name is Mohit Garg. He is the CEO and Co-founder of Oloid AI, an identification and authentication company using facial recognition. Mohit is a proven entrepreneur with 22 years of experience—building and scaling enterprise products and teams. Mohit was also an early member of successful Silicon Valley startups such as Wireless and Aruba Networks. He holds a master's degree from Stanford University, an MBA for the Indian School of Business and an undergraduate from DIT Delhi. Mohit, welcome to the show.

Mohit: Thank you Ari and Natalie. It's great to be on the show. Thank you for having me.

Ari: Yeah, Mohit, that's quite an impressive background. What I'd like to do is before we get into the company, I want to learn a little bit about yourself and your personal journey and how it prepared you for your entrepreneurial life that you live right now. Where did it all start?

Mohit: When I look back at my journey, I see three distinct phases, which led up to where I am today. It's always constantly changing so I’m not sure what we're doing next. But it started growing up, early years in India with my societal and parental pressure, which gave me only two options: to be an engineer or a doctor. And I was naturally inclined to be an engineer. I loved math and I loved science. So I spent pretty much my almost until 19 years of age in India going through the traditional education system. Did pretty well with academics. But I really didn’t have a sense of what it was leading to.

I got into engineering school because that's what the so-called bright kids did. But once I got into engineering school, I really started to connect with my passion for technology and connect with my passion for building products. I started to appreciate how technology is an amplifier, and how we were at the cusp of this big technology revolution, which is changing our lives. So I spent the first eight years of my life after my Master's in Electrical Engineering building products. Companies like Aruba Networks were very fortunate to be part of the early team in a company that grew so fast. So I got fascinated about building companies and not just products.

So in 2010, I founded my first startup. I founded it with three other friends from my business school and I really enjoyed the journey. Made a lot of mistakes along the way. And now as a second time founder and making new ones. So it's been a journey where I spent time as a developer, as an engineer, and spent time building the book of revenue in my previous startup as the Chief Revenue Officer. So it's been an interesting journey going from tech to revenue and now being a CEO.

I'm just looking forward to what the future has in store for me. It’s a journey of constant learning. That's how I like to describe it.

Natalie: So from your personal journey, how did you really come up with the idea? How was Oloid AI developed? What was your big aha moment?

Mohit: I call it progressive discovery, Natalie.

Natalie: Exactly.

Mohit: My first venture was in the sales productivity space. It's a very successful company named MindTickle. And as I was talking about what are my goals in my next venture, I want to align around the future of work. That is an area I was getting very interested in. And that's when I happened to meet my co-founder Madhu, who's our CTO, and he has a very deep background in AI identity management. He's worked on voice recognition AI 10 years back. He's got patents in Bluetooth identity protocols. So as we started to explore common interests, it was clear to us that computers with AI will have huge applications in the workplace.

So it wasn't like a right moment of we're going to solve this exact problem. It was more around computer vision for the enterprise is going to see huge benefits in certain parts of the enterprise in the workplace. We actually started with something very different. We started with an object recognition algorithm, where we set mobile safety applications where if you leave an object in front of the emergency door it will create alerts so that you don't have fire hazards if the emergency door is blocked. So that was actually the first MVP that we created, and then we found that it was a fairly limited market. So it's been a process of discovery and working very closely with early customers to align at the definition of the product Natalie. But I think it's about aligning with trends. It's about having the right team, and then working with early customers to align to the ultimate vision of the company.

Natalie: An evolution.

Mohit: Absolutely. Progressive discovery.

Ari: So once you finalized that the computer vision is the theme you want to explore with your startup, just curious, how did you go from object detection, your initial MVP, to facial recognition and the things you do today? What has that evolution, that journey been like?

Mohit: Yeah, I'll start by first clarifying that Oloid is not just a facial recognition platform. It is an identity platform. It is an authentication platform that uses several factors of identity. The face is one of them. We use QR code. We use voice. We also are including mobile based identity. So really what we're solving for is identity management and authentication for deskless workers.

But your question on how we arrived there is, as we were looking at implementing our object recognition, emergency door solutions and factories and warehouses, we were visiting these customers, early beta customers, and warehouses. And we were looking at how the day of the life of a worker goes about. And we could see that a factory worker or a construction worker has to rely on multiple methods of authentication. If they're using a time clock, or opening a door, or accessing a turnstile, or IoT device, sometimes they are using PIN pads by entering a code, sometimes fingerprints, the other time username passwords and guess what? They're carrying stuff. So entering using passwords is really hard.

So we found that identity management for the cognitive worker of the desk worker is a well solved problem. We all enjoy the benefit of singles. We have one username password, gives us access to CRM, email, third party applications. We actually even enjoined in our consumer lives where one Google password is giving us access to Google Drive, email, and a number of applications like Hangouts and whatnot. But for deskless workers, username and password is not a good solution. And from a security standpoint, if you give them badges, we realized that that was a security risk.

If I had your badge, I have your identity. So we got very convinced that a computer vision based contactless biometric, which uses face as a primary factor and you add additional factors to make it high confidence, high security, would be a great solution. We then started training our AI models around zero false positives to basically train it such that it gives you confidence score and you manage to a high confidence score on identity. Do fraud detection, spoof detection to ensure that nobody's spoofing masks or putting a picture in front of a camera. So this was a whole process of working on our initial inception of an idea. And then, I can talk about how we work with customers to even figure out the value, so we can go back, but it has to be valuable for the customer. So that's another second part of the discovery that we carried out after we had this idea about the technology.

Ari: Yeah, that's where I was going next. Your technology sounds pretty cool and nice to have, but it doesn't feel mandatory. So as I think about your go-to market, who was your beachhead market and customer that you felt strongly about that they could use your product and benefit from it?

Mohit: Because our initial anchoring of the value proposition was around enhanced security, in technical terms it's called identity assurance. When I say it is Person A, it is actually Person A, not Person B. In order to double click on the target segment market and the buyer, we had to double click on the drivers of ROI. So while removing friction for the end user and benefits like hygiene from a contactless technology was a great outcome. For the organization, we determined that identity assurance had sources of value. One was improved accountability and tech prevention and better compliance documentation.

In fact, if you go to a pharma warehouse who entered a certain goal needs to be documented because you may have storage of uncontrolled substances, right? So we found that high value industries, pharma and high value manufacturing would be a very good place for us to start. We also have aerospace and defense, but it was going to have long sales cycles. So we didn't pursue it at an early stage.

Second was the cost reduction because now you didn't need expensive hardware or some of the challenges with hardware breaking down due to contact based technologies. And lastly was payroll saving. So if you look at the inaccuracies of time in the attendance system because of impersonation, that can have double digit million dollars of savings. So we started with more of a security value proposition. And we penetrated highly reputed beta customers in high value manufacturing, pharma warehouses, and also construction. So that was our initial reach into the market.

Natalie: Thank you. Very interesting. And I wanted to talk about the challenges that you're experiencing. What challenges have you experienced getting to the market and what challenges are you experiencing today?

Mohit: This is something we constantly discussed within the founding team and also our entire team. We say hey, why is it that it’s still a source of risk and not eliminated because as you create value in a company, whether it's through funding rounds, the question you want to be asking is, what risks have you already eliminated? What is eliminated? And I think if I look at a few areas where we continue to make our investments, one is our education challenge.

Our customers have the option of retrofitting an incremental point solution. So let's take the example of what's happening with COVID-19. Before COVID-19, we were talking about primarily security, and it was a very strong value proposition. Today, customers care the most about safety, contactless interaction, contagion prevention at the workplace. So we'll talk about how we also have to do some product improvisations due to COVID-19.

Customers need this now. They need it today. So sometimes they have the option of using incremental retrofit technology, replacing a fingerprint reader with a badge reader and continue with business as usual. Our impact and success with a new generational shift of technology will be determined by the household and the market can appreciate the benefit of centralized identity. Use this COVID-19 as an opportunity to do a major transformation of identity. So that's an education challenge, which we constantly have to prove ROI and showcase that this is an opportunity to revamp your identity.

The second part of the education challenge is for the end users. So we all know that there's news around the unethical use of facial recognition by state agencies in a public context. This heightened concern around data privacy. The reality is that as part of our employment, we have always been entrusted with our employers with significant amounts of private data.

Employers have a picture, social security number, and whatnot. So when it comes to centralized identity management and use of contactless biometrics, we need to educate our end users that the data is used and retained only for the workplace applications. We encrypt the personally identifiable information with a one way hash to identify the data. So even if the employer wanted to provide biometrics to, let's say, state agency, the data is practically useless. It cannot be tied to an individual outside of proprietary software.

So there's two parts of the education, educating the market from a customer standpoint, from a user standpoint. And lastly, the third challenge, I would say is scaling up a company which has a physical product. So my first startup was a pure software application. It can be downloaded million times a day, whereas our software is to be downloaded into a tablet or phone. And sometimes the procurement of the device mounting and wiring of a device can result in some friction. So we are partnering with system integrators, installation service providers in the short term to address this, but we are depending on a pure self-install model that makes it very easy for an untrained technician to drive a few screws, install the iPad, connect to Wi-Fi, and get going. So I answered purely from a scaling standpoint. It’s the education challenge. And then the scaling challenge makes it very easy to deploy those activities we are working on.

Ari: Mohit, you're already touching on all the points I want to cover here. COVID-19, Congress regulating facial recognition, all the way to challenges of being a physical product company. So let's work backwards. Being a physical product company is hard. And it's a tough proposition even in Silicon Valley. I know some entrepreneurs who would give up on it too quickly. And so many VCs who would not touch it. So I want to know, what are some of the challenges you're experiencing in building and scaling a physical product company? And how are you navigating them?

Mohit: Yeah, that's a great point. And hopefully there are some learnings here from our journey for your listeners. We early on talk in a call that we want to be a hardware integrated product and not a hardware product. So the way I would describe it is we are not a hardware company. What we integrate with are off the shelf hardware. We do not manufacture hardware. We do not plan to manufacture. But we certify hardware that we can endorse and say that our customers can buy those pieces of hardware and the software will work with those without any issues.

So let me double click on this. Let's take the example of post COVID-19. We found that our customers needed to not just authenticate the identity of a worker-employee contractor visitor coming into the workplace. They also needed to put in some contagion prevention measures, check the temperature for elevated temperature. We have some QR code scanning with the self-identification questions. So the way we went about this illustrates our hardware integrated approach. We started testing off the shelf thermal cameras that were available in the market. We had started reading articles about employers in Asia and eastern parts of the world which saw the pandemic sooner than we did that they were using certain hardware for temperature scanning, contactless automated temperature scan.

We started sourcing these hardware as early as February this year. We’ve got these devices. We tested with more than 10 of these hardwares. We found fewer times to have the acceptable level of accuracy, which will be right for our customers. And then we provide integrations with them. Some of them are API integration, which are much easier to build. Some of them required some proprietary development. We continue to build Oloid as an integration platform. It can integrate with IP cameras. It can integrate with Bluetooth devices. It can integrate with QR codes. It can integrate now with thermal devices.

So we believe this is a very scalable approach. Our customers take advantage of Oloid software by downloading an app from the App store. The app comes free birth with these integrations so the customers can buy the hardware themselves. We do not have to supply them. But our customers do have to implement it in terms of putting the hardware on the wall if it's a tablet or a camera and installing and wiring it. So we do have that aspect of the physical implementation. But from a development standpoint, we continue to focus on being a scalable integration based software platform and all the hardware. I hope that answers the question.

Ari: Yes, it does. So you're taking more of a platform approach and not being a hardware company.

Mohit: Yes.

Ari: I like that a lot. Now, you touched on a lot of things here. And it seems like you're moving very, very quickly in terms of doing some of those integrations for the devices that are out there that measure temperature, and so on, especially in this global pandemic. So just curious, how are you managing it in terms of moving so quickly, with the solutions that you're coming up with while we're in a standstill mode?

Mohit: Yeah, I think that's why I give a lot of credit to our team. And as you can imagine, a startup, especially an early stage company, its success is hugely determined by the quality of the initial thing. And I think I speak for most startups and founders that I'm familiar with that software companies are very well designed for remote collaboration. We already had remote members in our team who are not even in our Sunnyvale office. We had close to eight members of the team. And now we have 10. We actually have two more hires that we made recently.

So we were designed as an organization to not rely on being in the same room. And for us, the collaboration post pandemic, I think there's only one challenge which was we lost access to our labs. So we are to set up labs in our garage. My co-founder has a lab setup in his garage. So I think there was this element of having to rebuild our lab setup. But other than that, from a remote collaboration standpoint, our team was able to flip the switch with no challenge. And I would say it is the mission, right? So today, because our solutions are enabling the Return to Work program for our customers, they are helping contagion prevention, our team is super excited..

We are part of the solution in these tough times. It's a time when the world is confronted with serious problems. And we are excited to be working with more and more businesses. We’re getting them back to work. So thermal scanning is just the beginning. We've already formed partnerships where we don't have to solve every piece of the puzzle. You have partners who are building IoT solutions for contact tracing and social distancing. We’re integrating with them. So we believe that we have a very strong identity solution and we partnered with hardware providers, IoT providers, to provide a comprehensive safety and security solution. And that's what keeps our team going. Every time we talk about the successes we are seeing with our customers, we all love to be part of this mission and collaborate remotely. It's been a time of exciting possibilities, where we think we're part of the series of challenges for our customers.

Ari: Yeah, that's definitely very exciting. Now turning to the market and sales side of it. What is your go-to market approach for COVID-19 tailored offerings? Because these are not normal times. So I'm just curious, how are you going to market with them?

Mohit: Yeah, so if you look at the global market, and we’ve sort of upgraded opportunities for our customers. So the easiest to implement as our solution is COVID-19 symptoms questionnaire, which means I'm a visitor or an employee coming to work. And every day, I have to answer 4 questions based on CDC guidelines. I have to attest that I'm not sick, somebody in my close contact is not sick, and I have not traveled internationally. There are customers who are doing it manually. We've gone back into a QR code scanning in the same camp as we use sufficient with application.

This is a free application for qualified businesses. Our go-to market is that - any business who wants to automate what they're doing manually, can download our app, and use our QR code system. They don't need any custom hardware. It works with any phone, Android, iOS. So that makes our entry funnel really broad. We don't even want to make money from all of these customers. Some of these customers are so small, restaurants and small businesses, we actually have no interest in making money from them. We just want to be of service. And then this creates the mindset of brand recognition and paid forward where these customers now will think about what is the additional value I can get on.

Then you can go for temperature scanning or thermal scanning, where they can buy a thermal camera, integrate with our platform, and this is going to be where qualified businesses can get into a subscription contract with us in the temperature screening. And at this point, they're still not taking advantage of old platforms but getting value. The third step is most of enterprises where our global market is really focused on where now you can bring in the whole identity framework, with thermal scanning QR code, plots into the identity framework, you can create the logic of using the QR code, I can just scan but I'm also using the identity to operate turnstiles, operate time clocks and now I have fully integrated solution.

So we have a layer up go-to market into these three stages where customers can opt into the fee offering, have inexpensive upsell and start using part of our platform and then have the experience of having trust in the technology to then bind to the full platform. So we're doing it both inbound and outbound. From an inbound standpoint, we're generating content and running online campaigns. From an outbound standpoint, it will target lists of customers, will be in a sweet spot. And we are reaching out to them directly through email and phone.

Ari: Great, great. And how big is your team and how the team is structured between sales and development?

Mohit: Yeah, so we are primarily an R&D focused team. Our go-to market team is pretty small. We're all a team of 20. And I would say that 70% of our team is all engineers and R&D people. We have a small customer success team. And a bulk of our lead volume is coming to channel partners and to our online campaigns. So at this point, it's a very small, highly focused sales team. Generation leads are coming from the website, and let us engage through showing the value of our product and closing deals.

Ari: Great. And the last point on the challenge, right? In the wake of these recent events, some larger companies have abandoned their facial recognition technology. Microsoft and Amazon are calling Congress to regulate it. And I know you touched on initially about keeping data private, and it's more for identification and authentication. So do you have an opinion on what's out there? What are your thoughts when we think about the clear view, and other companies that are out there that are providing, say police or to government officials with some of that data. There is a lot of debate and controversy around that. Just curious because you live and breathe in that space. What is your opinion?

Mohit: Yeah, so the first thing is that we want to be very focused on the workplace applications. So we have no intention whatsoever to provide surveillance technology for monitoring airports, traffic signals, basically public spaces. We have no intention of making our technology available in that area. And we are publicly very vocal about that. We are seeing a lot of traction in the workplace. And that's what we are passionate about. The second thing is even within the workplace, how can you bring a robust data privacy framework? So here, we have worked with worker unions to understand their concerns. We've worked with legal teams of our customers, and we have done our own privacy analysis of the data framework.

So we came out with some very foundational guiding principles. One is around transparency of how the end user is aware of their identity is being used. If an employee uses Oloid for time-clocking or for coping the go for the main entrance, the employee should have the ability to look at the log of how it is being used. The employer should not surreptitiously use it to track the bathroom breaks, for example. So we need all that transparency available to the end user. We also believe that as employees use this technology, they're going to enjoy the benefits and they're going to develop trust. So for us, the success will come from the employers and the employees. They use this technology, enjoys frictionless benefits, and it becomes second nature. Just like we don't want to carry the keys and insert them into our cars, the wireless keys do the job for us. Going back to physical keys is going to be hard.

But first, you have to learn to trust the technology. So both from a data privacy framework that I talked about, from a transparency that I talked about, we want to claim the ethical use of this technology, both facial recognition and non-facial recognition factors to provide a product that works for both employers and for the employees from a data protection standpoint. So that's an area we are very passionate about. And we are looking at publishing our frameworks as a best practice for maybe other vendors or companies to follow. We want to be part of that whole discussion and provide this framework, which I think is very powerful. It addresses all the concerns that have been highlighted in the debates. Those are more the public context than in the private workplace.

Ari: And now switching gears a bit. Being a Silicon Valley AI company to VC conversation is inevitable. A lot of the entrepreneurs that listen to this show are also thinking about it. And I know it's one of the biggest things on their mind. So given your experience building and scaling an AI company, what is your advice on raising capital?

Mohit: Yeah, so I've had the opportunity to have been through multiple fundraising rounds in my previous startup and now with this VC money, with Oloid as well. And through this experience, I’ve learned a few things. One is that it is not just about the idea. And I'm stating something which we care about a lot, but I'll touch upon the nuances here. I think it's about this understanding of how the VC model works. And being able to explain to the VC on how this idea matches with the VC model. Why is this a scalable company? Why is this for creating VC returns? And also, it is not a workflow company where you move x and bytes and you solve one problem. It's a lot about defensibility. It's a lot about network effects in technology.

And I find that many times as the excited and zealous entrepreneurs, we just focus on that workflow and we say our software can do ABC XYZ. But I think it's really important for a VC conversation to abstract on how that ABC XYZ continues to scale and becomes a massive transformation long term. How does it become a meaningful company? How does it remain within the company and not get substituted for a cheaper, faster product that comes along? So this is all rooted in how strategically you're thinking about your product? How does it become indispensable for the customer? How's it sticky? Why can't the customer just switch to a cheaper, faster product? I think this is one area where at least as a first time entrepreneur, I did it myself. I did not pay enough attention. And this time around, that's an active conversation that we have. It gives the VCs the confidence that this founding team, they are aware of how you can create VC scale returns, how you can create defensible technology? So I think this is a big part of learning along my journey of being very conscious and deliberate about defensibility with technologies, scalability of your tech.

Natalie: Thank you. And that's really interesting for us. And I did want to talk about the name of your company, Oloid AI. Can you just tell us a little bit more about how you developed the name? I did read about it on your website, and I found it really interesting. So I would love to hear a little bit more about the background of that.

Mohit: Yeah, certainly. Natalie, Oloid is a shape, which has very interesting geometrical properties and we were fascinated like I said, with a background in mathematics, our founding team, we were very fascinated about the shape. And we saw videos of that, and we thought it would be a very cool name for the company. But we also were building an identity company. And we wanted a name which had an ID part of it. In fact, we have a photo ID domain. And although it has a significance when it comes to some of the algorithms, there's a whole middle class algorithm which is relevant to us. And we only look once at the door when it opens. So for several reasons, we thought it would be a very appropriate name for our company. And our vision is that we'll have a company where our logo and this Oloid shape become synonymous with this technology of bringing to the workplace. Just excited about having a great brand and a meaningful name for the company to bring about this change.

Natalie: And I did see the little video of the Oloid on your website, and it was really soothing. Like it felt really nice. I'd like to think about the name a lot and I can see how that relates to your company as well and giving it a really nice image.

Mohit: Thank you. I have one at my desk, which is a good thing to play around with.

Natalie: Yeah. I keep watching it.

Ari: That's great. It seems like you're off to a good start here. And especially in these times, you're creating solutions that are much needed. But what is your vision? Where do you see the company, say, 3 years from now or 5 years from now?

Mohit: To build on few things I shared, which is how you build network effects, how do you build a product that is relevant long term, we believe that with our investments into making very robust technology, along with the privacy framework, we should be able to keep the trust of our end users, for us to commit and manage their identity on our platform, not just one employer, but across multiple employers. So think about the network effects where I have enjoyed the benefits of Oloid identity in employer 1, and now I have my preferences. I have my identities built into it, and I'm able to carry it. I'm a contractor with a gig economy, and I'm doing multiple gigs. How powerful it would be if I had my Oloid ID. And I'm able to access the devices and be able to access my work applications across multiple employers. We believe that we have to single sign on for the deskless worker. And for that, we are making all the foundational work around having a robust technology privacy framework. And we believe that what we do is keep the trust of our customers, keep the trust of our end users and the magic will happen.

Ari: And thinking about the buyers of your technology and the leaders that these industrial workspaces, what message do you have for them? How can they help you and themselves accelerate on this journey to a computer vision driven workplace.

Mohit: So when I look at the role of the chief information security officer, when I look at people who are responsible for logical identity, they end up being a different function than the function responsible for a physical identity that's managed by facilities or physical security. My message for the industry would be to promote the convergence of physical and logical security. This will have huge benefits. Today the data is siloed. The silos for the most part are managing their cognitive workforce username and passwords. And then managing the badges and the IDs for the pin pads and fingerprint systems. There needs to be a convergence between the two areas. This will provide much stronger security and provide agility to your business. Imagine the time spent today to onboard a new hire contractor, you actually have to spend a lot of money going to a test, printing of a badge, and then the data might not populate or flow into the right systems at the right time. So we believe technology has come a long way where this is feasible and possible. You have to take silos within the organization of being physical and logical security.

Natalie: Interesting. And before we wrap up, I just wanted to ask, do you have any advice for future or aspiring entrepreneurs?

Mohit: That's a very broad question. And I'm trying to think of distilling my comment into more actionable advice. I would say, one area where I would highly, highly recommend entrepreneurs is to think of their company and idea in two stages. One is brutal honesty and brutal validation of the idea and making that a dispassionate process. Once they have developed their own conviction, there is a database to stick with it. Every company, I've done it twice, and both of the times I found naysayers. I found skeptics. Once you have developed your own database conviction, it makes most sense to not switch your idea or change because some VC gave you some feedback that this won’t work. It really pays huge dividends to be contrarian sometimes. You just have to make sure that everybody's conviction was founded based on the data, based on your research that can stand by and work yourself convinced. It never serves well to keep on changing your pitch, according to every meeting. Sticking to your conviction is one of the strongest things to find the respect of your team and VCs in the long term.

Natalie: Thank you. That's really valuable, actually. So thank you.

Ari: Yeah, no, that's a deep advice, actually. So Mohit, thank you for taking the time out. It is Friday, and I know it's late. So really appreciate you taking the time out. This was super valuable, and super timely, given what we are going through in these times here. So really, really appreciate it. Thank you for being with us on our show.

Mohit: My pleasure. Amir, Natalie, it was great to chat with you.

Natalie: Thank you so much.



Ready to Get Started?