Suvrat Bhooshan

Founder & CEO Gan.ai

Last Reviewed: September 18, 2026

Executive Summary

Suvrat Bhooshan is the founder and CEO of Gan.ai, a generative-AI company built around a single deceptively hard idea: that a brand should be able to record one video and send a different, individually personalized version of it to every customer. The platform re-synthesizes the speaker’s voice and lip movements so that names, products, dates and other variables appear to have been spoken by the presenter, from as little as two minutes of training footage per speaker. Bhooshan came to it as a researcher rather than a marketer — he was working on deep learning over multimodal data at Facebook AI Research in New York, with published work at venues including NeurIPS and ACL, when he left in early 2021 to build the company. Gan.ai now serves enterprise marketing teams at companies including Samsung, Zomato, vivo and Mobile Premier League, alongside individual professionals such as realtors and SaaS sales representatives, and charges by volume of video generated. Forbes named Bhooshan to its 30 Under 30 Asia list for enterprise technology in 2022.

Career Highlights

• Named to the Forbes 30 Under 30 Asia list for 2022, in the Enterprise Technology category.
• Founded Gan.ai in 2021, after leaving Facebook AI Research in New York.
• Raised a $5.25m seed round announced in May 2023, led by Surge, the rapid scale-up programme run by Sequoia Capital India and Southeast Asia, with participation from Emergent Ventures and angel investors.
• Grew the platform to over 200 users including more than 40 enterprise customers, among them Samsung, Zomato, vivo and Mobile Premier League.
• Published research in deep learning on multimodal data classification, with papers at NeurIPS, ACL and other conferences, and released open-source deep-learning libraries.
• Extended the product beyond personalization into AI-powered video localization, automatically dubbing a recorded video into additional languages.

Professional Journey

Bhooshan trained as a computer scientist, taking an undergraduate degree at Georgia Tech and a master’s at Stanford, and moved into deep-learning research rather than product work. Earlier in his career he worked on artificial intelligence at the Unique Identification Authority of India, the body behind the Aadhaar identity system — an unusually large-scale environment for applied machine learning. He then joined Facebook AI Research in New York, where his published work concentrated on deep learning over multimodal data, the combination of text, audio and video signals that would later underpin his company.

He left FAIR in early 2021 to start what became Gan.ai. The technical premise was that generative models had reached the point where a human presenter could be convincingly re-synthesized at the level of individual words, and the commercial premise was that marketing video had remained stubbornly one-to-many while almost every other channel had become personalized. The product asks a brand to record a video once and mark the variable parts of the script; the system then generates a distinct rendering per recipient, matching both voice and lip movement, at a scale of thousands or millions of copies.

Recognition arrived quickly. Forbes placed him on its 30 Under 30 Asia list for 2022 in the enterprise technology category, and Gan.ai was among the technologies selected to exhibit at TechCrunch Disrupt that year. In May 2023 the company announced a $5.25m seed round led by Surge, Sequoia Capital India’s scale-up programme, with Emergent Ventures and angels participating. At the time, Bhooshan described the roadmap in terms of latency as much as quality — making generation instant and real time — and said the money would go towards a US go-to-market team and towards recruiting more consumer brands.

Since then the company has pushed in two directions at once: down, into the amount of source material needed, having reduced per-speaker training data to around two minutes; and sideways, into localization, using the same voice and lip-sync machinery to dub existing video into other languages. Bhooshan has become a regular commentator in Indian marketing and technology press on where video personalization is heading, arguing that it will shift from a differentiator to a baseline expectation for brands.

Education

Stanford University

Master of Science, Computer Science

Georgia Institute of Technology

Bachelor of Science, Computer Science

Forbes confirms degrees from both institutions; the specific degree levels and fields above come from secondary coverage rather than a primary record.