At The Tech Factor by Roopa Kumar. The podcast was an interesting conversation between Gopi Prashanth, Founder at Fithub and Roopa Kumar, Founder & Group CEO, Purple Quarter. In conversation with Roopa, the Ex-Amazon techie turned Entrepreneur talks about revolutionising fitness through Gen AI technology. As a serial investor, Gopi touches upon the India investment sentiment, his resonance with Bezos’ Regret Minimization Framework and his decision to pivot into the fitness industry powered with AI.
Gopi Shares,
Before we zoned in on computer vision, we built and killed 18 technologies—18 different ways of understanding what a user did in the store. These were partially computer vision and non computer vision that were all built and destroyed in our labs. We had to figure out and learn every single technology possible, such as RFID stickers, and quickly experiment with it.
The question wasn’t just whether it could solve the problem today; it was also whether this has a path - algorithmic, operational, or hardware—so that we get to make a high fidelity transactional system out of it. When you consider the raw signals coming in from all the different sensors that we ended up using, we came to the conclusion that computer vision was the most malleable signal that we could conquer.
Because your algorithms could continuously evolve over time without having to go back and change the hardware itself, it could take us closer to where we wanted to be for transactional fidelity, which is why we zoned in on computer vision.
If we look at the Palm payment software at Amazon, it came out of a need wherein computer vision could be used to make a human palm their identity because there is enough unique signal over there. They do not require a credit card or a phone. They themselves became their biometric sensor and it was based on very similar computer vision as a malleable platform, on top of which we can keep improving the algorithms to get to the transactional fidelity that is needed for almost credit card payment level accuracy.
In the context of the regret minimization framework, I identified two significant regrets. The first revolves around entrepreneurship, and the second centers on neglecting my physical fitness and mental well-being. To put it simply, these two aspects were my paramount desires for correction.
Before I realised, I had become too preoccupied with wealth creation, career growth (becoming a director at Amazon, long hours coding in the gaming company) and worked towards increasing salaries for a better lifestyle. But it's an endless cycle; once you get in, it's never enough.
So in that process, the thing that basically ended up falling by the wayside was my personal health and wellbeing along with mental well being. I didn’t experience any episode as such, but simply coming to terms with the fact that I was not making that a top priority, was eye-opening.
And when I started thinking about my fitness journey, I realized that all the apps out there are just one-size-fits-all. I needed something that was molded around me, that understood me and was able to adapt to what I wanted every day.
I didn't want to go and do the research to acquire the knowledge needed to be my own trainer. But at the same time, finding a trainer to do this for you on a consistent, daily basis is a very expensive proposition. So when I saw this problem, that's when I realized Generative AI is the perfect solution for this because it can hyper customize for the long tail of use cases. We have figured out the magic behind how to make Generative AI hyper customize things for us. And now it's very quick for us to keep adding new skills or new AI trainers to the platform.
The AI trainer is a chatbot. But unlike a chatbot, when you turn things off, it doesn't forget you. It remembers who you are. And every single thing that you have said, is condensed into things that matter.
Watch the full podcast here,
We place CTOs, CPOs and senior technology leaders for high-growth companies.
Occasional, high-signal analysis on hiring and leading technology teams.
Read next
