Tell me about your role at Upside AI and what it entails?

I lead Sales at Upside AI. This means I am responsible for adding new clients, channel partners and distributors. Given that we are creating a category, a lot of my pitches are around evangelizing AI in investing, talking through why a rules-based process is superior to human intuition. We have grown 10x over the last year so we have been hard at work! I am also responsible for existing client relationships – talking through portfolios, performance, etc.

You moved from a career in financial services to one in fintech. How has the transition been to absorb and understand the tech aspect of your current job?

Given that I am a CA and a CFA by training and have been in finance and investing my whole career, the “fin” part of fintech came intuitively. Further, when I was in venture capital, we evaluate early stage startups and tech is at the core of anything new that is being built. I’ve always been technologically inclined – during the pandemic last year, I did some courses to learn Python so I can atleast read part of the code we write at Upside AI! Given that we were building the products from scratch, the machine learning logic was something Nikhil, Atanuu and I had to come up with together. So those months of iterations teach you not just your own product, but also make you more comfortable with deeper tech involvement.

What are the biggest myths that currently exist about female founders of tech companies?

I think most female founders will tell you they are just founders first. Their gender is secondary/ irrelevant. But outside in, there are situations where the assumption is I am not as well versed with my tech as Nikhil and Atanuu. But again, I believe that’s because I am on in sales. If I was an engineer or headed product, I would not face these biases. Therefore, I find that any “myths” relate more to my role definition and not my gender.

What challenges do women face in building a career in emerging technologies like AI? How can the business community and society address this issue collectively and efficiently?

I’ve thought about this a lot – studies show that generally interest in STEM till school is reasonably equal between boys and girls. However, the skew in STEM careers weighs heavily to men. Social conditioning around how STEM is not a “feminine” subject, more boys in a field organically leads to boys’ clubs which ends up excluding the few women who join at entry level. I think, until school, support for boys and girls in STEM is broadly equal – the drop off in support is steep thereafter. We need to encourage more girls at college/ graduation level for starters. Women mentors for other women at entry level jobs intuitively sounds like it would help. Having said that, there are other studies that show that while boys and girls have similar aptitude for science and math, girls are relatively better at subjects like reading. So, even after correcting for societal biases, it may also be that more women are interested in (and better than) the arts, relatively.

What is your biggest AI nightmare?

When we build products, we are most paranoid about (1) over/ underfitting and (2) data cleanliness. There is a saying that “if you torture data long enough, it will confess to anything”. We do very extensive tests to ensure we are not overfitting – out of sample testing, purposely including spurious data, running diagnostics on the right balance of iterations to run, etc are a core part of building the product. Secondly, the sanctity of data is paramount. A lot of our time also goes into data clean up and running tests to make sure our data is clean and complete. 

Also Read: Startup of the Week: The ‘upside’ of using AI to manage and multiply wealth