Sravanth
Aluru

There is a certain kind of person who cannot leave a good thing alone. Sravanth Aluru is one of them. He had a comfortable seat at Microsoft in the early 2000s, running programs and quietly noticing that computers were starting to do something remarkable - they were beginning to recognize what they were looking at. Most people would have filed that observation away. He kept it.

Then came the detour that reads like a plot twist. Aluru went to Wharton, earned his MBA, and landed on Wall Street, where he became a technology investment banker. Not a small one, either. He rose to head technology for Internet and Digital in investment banking at Deutsche Bank and advised on tech transactions worth, by his own accounting, more than $100 billion combined - mergers, acquisitions, and the kind of Nasdaq IPOs that make careers. He was, by every conventional measure, a success at a game most people would happily play for life.

He quit to build a company about pixels.

In 2014, well before generative AI was a phrase anyone repeated at dinner parties, Aluru founded Avataar. The pitch, if you can call an engineer's private conviction a pitch, was that the confluence of self-learning AI and computer vision would change how humans see everything they buy, browse, and touch online. This was a wildly early bet. The technology was not ready. The market was not asking. He started anyway, which is either foolishness or conviction, and the difference, as usual, only becomes clear in hindsight.

Aluru holds an engineering degree from IIT Bombay, that famously demanding forge of Indian technical talent, and he is a lead inventor on more than ten US patents spanning AI, computer vision, and the unglamorous but essential art of moving data efficiently over mobile networks. He describes himself, disarmingly, as "a technology enthusiast and a student for life." Coming from someone with his resume, that could sound like false modesty. It reads instead like a working philosophy.

For years Avataar did the patient, unsexy work of turning ordinary product photos and videos into life-size, interactive 3D models - at scale, which is the part that matters. Samsung came calling. So did the retail giants. The company built a first-of-its-kind platform that let shoppers spin, zoom, and inspect a sofa or a sneaker as though it were sitting in their living room. It was clever. But cleverness alone does not build a business.

The turning point arrived in 2019, and it arrived the way good science does - through a test. Avataar ran an A/B comparison: flat 2D product images against interactive 3D. The result was not subtle. Brands using 3D sold roughly three times more. That is the sort of number that ends arguments. It handed Aluru something founders rarely get so cleanly - a single experiment that validated the whole thesis of the company.

What separates Aluru from the crowd of AI founders now flooding the field is where he chose to point his ambition. Text generation, he is quick to note, is largely solved. Image generation, too. So he went after the hard problem. "At Avataar, we take it a step further, extending generative AI to the 3D world," he has said. The distinction is not marketing. Generating a believable 3D object means the machine has to understand physics - how light falls, how a matte surface differs from a glossy one, how a reflection behaves. It is the difference between painting a picture of a glass and understanding what glass is.

By 2025 the company had pivoted its energy toward video, launching a tool called Velocity that automatically generates product videos from nothing more than a product link. The economics are almost rude. Where a polished product video once cost thousands of dollars and weeks of production, Avataar's Varya platform can produce a 211-second video for roughly a hundred rupees - a little more than a dollar. When the cost of something falls by two orders of magnitude, the thing itself stops being a luxury and becomes a default.

Aluru is careful, though, not to confuse motion with meaning. He has grown impatient with vanity metrics. "The scale of evidence is growing, with transactional outcomes now becoming evident, not just engagement levels," he says, and then, ever the banker-turned-engineer, he reaches for the receipt: a 6.7 percent conversion lift at Sleep Number. Not clicks. Not dwell time. Sales.

Today Avataar runs on two clocks - Silicon Valley and Bengaluru - with a team of around 180 building for customers including HP, Victoria's Secret, Lowe's, Newegg, TVS, and Bajaj. The funding has followed the conviction: more than $55 million, including a $45 million Series B led by Tiger Global with Sequoia Capital, plus backing from Peak XV Partners. Aluru's aspiration now stretches past the shopping cart. He talks about spatial storytelling spilling into gaming, real estate, and industrial applications through an open SDK - the belief, put plainly, that enterprise and consumer interactions will all eventually go 3D.

It is a big claim. But then, the man made it in 2014 and simply waited for the world to catch up. Patience, it turns out, is just conviction wearing a watch.