Debashis Nayak

• Owned the auto insurance vertical end to end as a $9.6M ARR business unit ($960K/month), split 42% B2C in-app, 51% B2B partner distribution, and 7% extended warranty. Promoted twice in 18 months.
• Grew revenue per click from $7 to $20, a 192% lift, by rebuilding personas from behavioral data on Amplitude and Snowflake and stacking signal-based enrichment instead of routing on declared intent.
• Pitched, scoped, and shipped an insurance embed inside Microsoft Copilot chat, working directly with the Copilot product and engineering teams. Formal partnership signed, live in beta; SDK designed to port to other AI assistants.
• Built the B2B distribution layer: partner pages, embedded comparison widgets across 15+ high-traffic sites, and API contracts surfacing 58+ insurance brands to embedded-finance partners, neobanks, and parent-company portals.
• Designed the Vehicle Scoring System (a credit-score-for-cars on hybrid rule and ML over telematics and behavioral signals) and Garage, the in-app vehicle vault that feeds it.
• Led product planning for the Trellis acquisition (US insurance license) and a team of 8 engineers, 1 designer, 2 data scientists, and 1 direct-report PM across the US and KL. Offered Director of Product post-acquisition; declined the US relocation.

• Built and delivered the PM training curriculum end to end, covering product strategy, user research, roadmapping, Agile, and data-driven experimentation. Achieved 95%+ job placement rate across cohorts.
• Coached 50+ PMs into product roles. Taught funnel analysis, A/B testing, stakeholder management, and metrics-driven decision making through real product cases.

• Defined and ran product strategy for a US healthcare SaaS line. Translated complex clinical workflows into requirements that aligned with customer needs and could be shipped within the contract timeline.
• Led a cross-functional team of 10+ across product, QA, and UX in an Agile cadence. Hit every product launch milestone on time and within the contract budget.

• Owned the merchant payments product on an RBI-regulated platform with 10,000+ active merchants. Defined API requirements, payment gateway integrations, and compliance workflows.
• Shipped Split Calculator from zero to production in 3 months with 120+ merchants live. The first bill-splitting and shared-payments product on the platform, with adoption across both small merchant and consumer segments.
• Hired and scaled a team of 8 across product, QA, and UI/UX. Built the event-tracking framework that cut feature instrumentation time by around 20% and made data-driven prioritisation possible across the platform for the first time.

• Led product development for ePOD, a digital delivery verification product for fleet operators. Cut delivery disputes by 8% and lifted last-mile accountability across enterprise fleet customers and shipped Control Center, a real-time shipment-visibility platform tracking live deliveries across enterprise fleets.
• Owned the product roadmap for a logistics SaaS platform with 15+ enterprise clients, integrating 3PL partners and fleet management systems into the customer dashboard.

• Founded CoolHax Labs out of college and scaled to 50+ people. Shipped 100+ B2B and B2C products across India and international markets. Generated ₹5 crore in revenue over eight years with no outside funding.
• Built hardware-software products including smart mirrors and IoT-enabled hospitality devices for enterprise clients across retail, hospitality, and consumer electronics. Held 90%+ retention across 100+ accounts through fast iteration and hands-on customer discovery.

Open-sourced my personal AI operating system as Holt, a local-first framework for running AI agents against your own files. Shipped to npm and Homebrew with per-folder semantic memory, pluggable skills, an MCP server, and scheduled background agents. Built end to end by orchestrating parallel AI coding agents that I spec, review, and merge. 5,200+ npm downloads since the July 2026 launch, now at v0.18. Now building Holt Teams, a team edition where shared AI memory stays on the firm's own machines.

Designed, shipped, and monetized an AI Vedic astrology platform, live in India with ~300 users and its first paying customers. Backend on Hetzner with Coolify, frontend on Vercel. Architected the full system from Mongo schema and content pipelines to the design system along with AI orchestration.

Building a B2C AI resume, Job Search, with interview-prep, currently in beta with early paying customers. Full-stack on Next.js, MongoDB, Clerk auth, and OpenAI plus Claude APIs. Self-funded, self-built end to end.