Director, AI Product Lead - Zee Entertainment Enterprises Ltd. - Delhi NCR, India
(2026-03)
- Took Trinetra from zero to ₹2 Cr (~$230K) in enterprise revenue within 90 days - no existing product, team or customer base. Owned P&L, pricing, model portfolio, commercial terms and go-to-market, and led 15 direct reports across engineering, ML, design and content production.
- Scaled AI content production to 6-7 finished minutes per day and 3 complete microdramas per month - script to dubbed, lip-synced video at ₹12,500 per finished minute, a 4-8x cost reduction against traditional production.
- Found that raw model acceptance ran 20-40%, making retakes rather than generation price the core cost driver. Shipped RAG-based prompt rewriting, storyboard-first approval and prompt caching (90% token cost reduction), shifting the goal from cheaper generation to right on the first attempt.
- Signed Zee's content division as anchor account, negotiating IPR assignment, confidentiality and content-access governance directly with legal, then extended the model to independent studios and won a pilot with Big FM (60+ national stations).
- Defined the core product, Vidya - a conversational agent where creators describe a scene in natural language and the system proposes studio-grade prompts, then generates video with dialogue, music and lip-sync - replacing a fragmented five-tool workflow with one surface, including a browser-based dubbing pipeline built after partner interviews identified dubbing as the highest-friction step.
- Architected a multi-provider strategy across 11 providers and 15+ models with priority-based fallback routing and config-driven rollout, absorbing two provider outages with zero user impact. Built the evaluation framework scoring quality, latency, cost per generation, policy reliability and language support that drove provider selection.
- Designed multi-tier usage-based pricing across creator, agency, enterprise localization and reseller segments, decoupling customer price from volatile vendor cost. Defined content safety and governance: automated moderation, real-person detection and copyright-safe music policy.
- Shipped two adjacent products - an in-house music generation tool replacing licensed music APIs, and an AI script analysis product producing ROI, casting and budget forecasts with automated quality gates.
Product Manager - Aftershoot - Noida, India
(2023-07 - 2026-03)
- Led a team of 5 product managers and APMs, owning strategy, roadmap and P&L for Culling, Editing and Retouching - the product lines that took ARR from $3M to $13M (4.3x) across 33K+ professional photographers globally.
- Lifted trial-to-paid conversion 39% by rebuilding the Culling and Editing workflows on the back of 100+ user interviews: simpler journeys, contextual guidance, less UI.
- Launched Creator Marketplace 0-to-1 in 8 weeks with 10 creators and 30+ edit styles; lifted NPS from 63 to 89 and Editing-funnel conversion 43%.
- Cut churn 27% via cohort-targeted win-back programs for 30- and 60-day inactive users, targeted by plan, geography and usage pattern.
- Worked with the ML team to optimize on-device inference for culling models, improving model performance 32% and cutting compute cost $150K a year.
- Shipped OTA update infrastructure, taking latest-version adoption from 37% to 72%; took Tethering from concept to beta in 8 weeks, acquiring 3K+ early adopters.
- Ran delivery across a 15-person cross-functional group spanning engineering, ML, design and marketing in two time zones.
Technical Product Manager - Monotype Imaging - Noida, India
(2022-03 - 2023-06)
- Owned Search and WhatTheFont, the ML-powered font identification tool, serving millions of discovery queries across web, desktop and enterprise products.
- Led the roadmap for AI-driven relevance and image-based identification with ML, infrastructure, design and licensing teams, improving accuracy 25-30% and cutting time to discovery 40%.
- Drove 24% uplift in font-discovery-to-licensing conversion, raised NPS from 55 to 90, and built Search as a Service on Algolia - a shared platform adopted across several internal and customer-facing products.
Product Manager - Talentopedia - Gurugram, India
(2019-01 - 2022-03)
- Joined as the first product hire, promoted from engineering, and built the product function from scratch - hiring and managing front-end, back-end and QA engineers and establishing the development process.
- Launched web and mobile 0-to-1, scaling to 8K DAU and 22K MAU at 4.8★ through product-led growth with minimal paid marketing.
- Built a lead management engine generating ₹80L revenue in its first two months; grew organic traffic 47%.
- Improved retention 30% year on year and cut onboarding drop-off 18%, driven by 50+ user interviews.
Front-End Developer - Talentopedia & Hammock Ventures - Gurugram, India
(2017-08 - 2019-01)
- Built high-performance platforms (40% faster load times, 26-40% traffic growth) and booking and payment systems that lifted conversion 18%.
