Deepak Prasad — Senior Product Manager

Senior Product Manager with 7+ years of experience across subscriptions, B2B marketplaces, and applied AI. Owned the subscription business through €659K in FY2025 revenue, built payout systems moving ₹20–25 Cr a month, and shipped an in-app AI coach that grew from ~300 to a peak of ~4,500 DAU after its November 2025 launch. Available to join immediately · Bengaluru.

Work Experience

Senior Product Manager — Sportstech

Oct 2024 – Sep 2026 | via Tejmonvi Softwares (ODC) · Full-time · 174,000+ Users · €659K FY25 Revenue

Owned end-to-end subscription strategy, consumer AI initiatives, and connected product experiences for a digital fitness platform across iOS and Android.

Product Manager — Sportstech

Aug 2023 – Sep 2024 | Full-time · 0→1 Platform Build

Joined as the first Product Manager on the platform, establishing 0→1 product foundations, user onboarding, and monetization architecture.

Product Manager — ReshaMandi

Jun 2021 – Sep 2023 | Full-time · B2B Agri-Tech Marketplace · ₹2,000 Cr Platform

Owned core product delivery across India’s sericulture value chain spanning ~1.1 Lakh stakeholders across 5 verticals (including 80K+ farmers via ReshaFarms).

Associate Product Manager — LionCircuits

Jul 2018 – May 2020 | Full-time · IoT & PCB Manufacturing Platform

Led concept-to-launch of a B2B Assembly Ordering Platform, contributing to a 40% increase in monthly orders.

Digitising a Complex B2B Marketplace — ReshaMandi

Understanding ground-level silk trading workflows and designing connected digital products across farmers, collection centres, payments, and pricing.

Replaced ad-hoc paper notes and WhatsApp handoffs with a structured digital lot record and a governed payout flow (<₹5L settled instantly, >₹5L within 2 hours at 99.9% success), contributing to scaling monthly marketplace volume from ~₹10–15 Cr to ₹20–25 Cr across 80,000+ farmers on a ₹2,000 Cr platform. Piloted image-based price recommendations with human-in-the-loop across 2 centres, and added geofenced bidding with timed payment windows.

Building a Conversational AI Coach — Sportstech

Taking an ambiguous "add AI" idea from concept to production: defining what to build, where to set boundaries, how we chose the models, and what we learned when real users started testing it.

Spearheaded 0→1 development of conversational in-app AI Coach (Gemini primary, ChatGPT fallback); scaled from ~300 to a peak of ~4,500 DAU after its November 2025 launch behind physiological safety guardrails.

Turning a Free Fitness App into a Subscription Business — Sportstech

Designing the end-to-end monetization experience: from packaging and paywalls to onboarding, value realization, conversion, and retention.

Co-authored the FY2025 subscription analysis setting monthly-to-annual migration as the 2026 priority. Supported growing the subscription engine to 12,401 paid subscribers and €659K FY2025 revenue at 81.9% YoY growth and 96.8% yearly-plan retention.

One Body. One Score. One Ecosystem. (Product Strategy & PRD)

A development-ready product strategy and cross-platform PRD uniting a fragmented connected-fitness ecosystem under a single progress model (unlaunched product strategy & PRD specification).

Authored the cross-platform PRD, specified the 0–100 Athletic Reliability scoring algorithm, and designed the 3-phase data migration strategy across 5 engineering surfaces (iOS, Android, Display, Smart Gym, Firmware).

Scaling Content Localization with AI — Sportstech

Redesigning a traditional video-production workflow into an AI-assisted localization pipeline across three European languages.

Shipped 200+ high-production workout videos across Italian, French, and Spanish in roughly 3 weeks (~10× faster than studio benchmarks), creating a repeatable operational blueprint for rapid market expansion.

Behind the AI Copilot: Transparent In-Memory RAG

An architectural deep dive into why RAG beats fine-tuning for domain portfolios, why in-memory cosine beats external vector databases at this scale, semantic chunking, and safety-first confidence gating.

Built with fast CPU in-memory cosine retrieval, zero external database dependencies, 100% citation grounding, and 95% (19 of 20) on a golden evaluation set.

Contact

Email: shipwithdeepak@gmail.com

LinkedIn: linkedin.com/in/prasad-deepak

GitHub: github.com/shipwithdeepak-code