Apply Now

Leveraging Big Data: from Payments to Food Product Innovation

Posted:
5 November, 2021

“In the end, we are our choices.”

This famous quote by Jeff Bezos at Princeton University resonated deeply with Himanshu Upreti. Years on, it has become one of his driving forces when making life’s pivotal decisions.

He aspired to make a global impact. And he knew that to do it, he had to solve immense, structural problems. 

What follows is Himanshu’s tactical gutsy journey from driving early impact with big data in the payments arena to founding Ai Palette at Entrepreneur First, to now, where his company successfully raised its Series A to further its roadmap developing data-driven tools for Consumer Packaged Goods (“CPG”) product innovation.

 

A Timely Discovery

His early quest searching for industrial applications for his mathematics and computing skills opened his eyes to the then-evolving world of big data and machine learning. 

Amid the explosion of available digital data, it was becoming increasingly evident for corporations of the need to utilise data science technologies to clear the noise to support critical business decision-making. 

“The future of humanity is certainly going to be integrated with big data,” thought Himanshu, “and I’m standing right at the precipice of that seismic change.” 

With that revelation, Himanshu chose to join payment leader Visa straight out of college. 

There he dabbled and honed his big data and machine learning skills, experimenting and building cutting-edge artificial intelligence models from zero to one while enjoying the thrills of problem-solving fresh daily challenges as competitors tried to catch up.

He also pioneered a Queue Management Tool that drove a more than 30% improvement in operational efficiency for the Visa Big Data Platform and experienced the joys and satisfaction of building an impactful data science product from scratch. 

These milestones became the hallmark of his three-year stint at Visa. 

So intrigued was he by the massive value creation data science brings to the payment field that he started musing about the expansive possibilities this would mean for a myriad of more conservative industries including food, oil & gas, and manufacturing. 

This intense curiosity reminded him once again of his ambition of solving big-world challenges.

He might have achieved much at Visa and was on a fast-paced, upward trajectory, but fintech was fast becoming commonplace. If he wanted to ride on his quick momentum at an accelerating rate, he needed a bigger problem in a still-traditional space – that’s where there will be the largest value add with data science. 

And he had to decide on that ‘mystical’ industry fast – being the earliest entrant in an old fashion market yields the highest return for his quality and robust machine learning solutions.

His only roadblock: he knew no one outside of the finance and information technology domains. 

When he chanced upon Entrepreneur First, it was as if the door to a new world of untold possibilities swung open. 

“Immediately connecting with a pool of exceptional individuals from diverse fields was exactly what I needed to kickstart my entrepreneurial dream,” recalls Himanshu, “I have to jump on this opportunity and so leaving Visa to join Entrepreneur First was a straightforward decision for me.” 

"Putting all your eggs in one basket might just work out if you have control of what is in it"

The freedom to dictate his trajectory motivated him far more than the fear of the unknown. The stakes might be high, but with great risk comes great reward.

From Day 1 at Entrepreneur First, Himanshu was fixated on achieving two agendas: going all-in on his technical edge on predictive analytics and finding the ‘high potential’ industry to apply it to.

One of his earliest and fondest memories of the platform was when fellow cohort member Somsubhra GanChoudhuri gave a talk on the Fast Moving Consumer Goods Industry (“FMCG”). 

As a foodie, with a self-professed obsession with healthy eating, he was naturally drawn to the idea of new food and drink product research and the science and art behind designing that mouth-watering, emotion-inducing flavour that makes eating and drinking pleasurable for many. 

“It was interesting to see, for example, how an innate understanding of distinct coffee bean nuances could gel well with precise experimentation techniques to create unusual coffee combinations,” explains Himanshu.

“We need to know that Brazilian beans give a great mouthfeel, Mexican coffee boasts of a good aftertaste before we can mix and optimise for exotic combinations like Kenyan coffee, Sumatra coffee, and more.

And these innovations will then appeal to seasoned coffee drinkers who are after that complete and balanced cup that comes with the right blend of Brazil and Mexican beans.” 

The big eureka moment for Himanshu came soon after when Som followed on and shared a staggering, mind-blowing statistic: for every 10 new food product innovations, 9 are destined to fail within the first year, costing food brands US$40 billion annually worldwide.

“When something is important enough, you persist even if the odds are not in your favour”

FMCG giants face a huge dilemma: either continually decode food trends and find that winning formula to set them apart or lose market share. 

They simply have to plow in cash for food product innovations even if the burn rate reaches near uncontrollable levels.

“I instinctively saw potential there,” reflects Himanshu “If we could just lower down the failure rate by say 10-20%, we would have already won.” 

And building a predictive analytics platform that turns organic online data into actionable insights for FMCG brands was the most obvious way.

The Birth of Ai Palette

Given that data to the machines are all but the same, Himanshu was sure his experience combining and processing previously siloed non-numerical data at Visa and teaching the computer to identify payment trends could be easily transferred to poring through macroscale data and scouting for the latest food trends. 

He convinced Som that they complemented each other – him on the technical front and Som on the domain front – and they founded Ai Palette. 

“Food and culture are interwoven,” says Himanshu. “There’s a cultural significance in the way food is prepared and served. So, we put our minds together to build a platform encompassing predictive analytics, natural language processing, and image recognition capabilities to take into account the nuances in languages of different cultures when making sense of organic online data. 

By doing so, we can enable global CPG brands to craft and launch their next-best selling products in different locations.” 

Ai Palette has since grown by leaps and bounds, and in three short years, they have expanded and now serve brands across Southeast Asia, China, India, Japan, the United States, and Europe.

How useful was this post?

Click on a star to rate it!