Prem Dutta

I build AI products. Specifically, I take generative AI from an interesting demo to something customers pay for, which mostly means owning the unglamorous parts: model selection, inference cost, latency, evaluation, and pricing that still works when vendor costs move under you. I have been in product for seven and a half years, and in technology for nine. Most recently I was Director of AI Product at Zee Entertainment, leading a team of 15 across engineering, ML, design and content production. Before that I led product at Aftershoot, a computer vision startup, and at Monotype Imaging I built ML-powered search and identification products. Two things I am proud of. At Aftershoot I owned three product lines through growth from $3M to $13M in ARR, a 4.3x increase, serving 33,000+ paying customers across 50+ countries. I lifted trial-to-paid conversion 39%, cut churn 27%, and cut on-device inference cost by $150K a year. At Zee I built Trinetra from nothing to a revenue-generating business inside 90 days, with Zee's own content division signed as the anchor enterprise account. I architected it as an orchestration layer across 11 model providers so we could swap or add a model through configuration rather than code, and I carried the P&L, including cost per generation and margin on every transaction. What I want next is a senior or lead product role at a company building AI products where the hard problems are real, fully remote or hybrid from Delhi NCR. I work best on 0-to-1 and early scaling, where the question is still whether the thing should exist and what it should cost. I am hands-on by preference, I prototype in code before I write a spec, and I would rather own a narrow problem completely than supervise a broad one. I am out of my notice period and available to start immediately.