Headline

AI Technical Product Manager @ Sigmoid | 13 yrs - Engineering & Product | I build v0 AI products myself, then scale with eng | 10+ AI data products GTM for Fortune-100 CPG Pharma | Agentic · RAG · LangGraph · Claude Code

About

I build AI products the way engineers build software — by shipping working prototypes, not slide decks. As a Technical Product Manager at Sigmoid, I lead full-stack, AI-powered data products for enterprise clients across CPG and Pharma. For Reckitt, a data-product marketplace, planogram tools and target/incentive/sales-hierarchy management tools; for Pfizer, an AI-driven visit-pitch planner for medical representatives; for PepsiCo, logistics tooling for truck-load optimisation; for Kenvue, a data-product marketplace with data-sufficiency modules. What makes 'how I work' different: I prototype the UI in Claude Design and ship v0 builds with Claude Code(schema, architecture, performance and security defined development), delivering roughly 30% of the working product myself before handing off to engineering for modularization and scale - tighter feedback loops, fewer misread requirements, and a product vision already pressure-tested in code before a single sprint is planned. Since I've built for enterprises throughout, security and data-privacy flows are designed in from the start - not bolted on later - so every AI-first product I ship is built to clear architecture review board (ARB) and technical review board (TRB) panels. Outside client work, I build to stay sharp on what's actually shippable: • Jarvis - a Manifest V3 Chrome extension that drafts smart replies across Google Chat and Gmail, using prompt-caching and a Haiku→Sonnet cascade-routing setup to keep latency and cost down. • TEO - a LangGraph multi-agent QA orchestrator with human-in-the-loop gates and 3-tier model routing. • Bharat AI Capital — an equity-research dashboard for Indian AI-infrastructure companies, with an LLM-driven analysis panel. Underneath all of this is a decade plus experience leading QA and quality engineering across health-tech and enterprise software. That background is my edge as an AI PM: I think reliability-first. I instinctively ask how a feature fails, how you'd test a non-deterministic model in production, and where the human gates need to be - questions most PMs reach for too late. Tools & methods I work in daily: Claude Code, Claude Design, Lovable, Cursor, Replit, n8n, RAG, Agentic Workflows, Multi-agent systems, LangGraph, MCP, Prompt Engineering, and Automation. Open to senior PM / AI-PM roles. Always happy to talk shop on agentic systems and prototype-driven product development — reach me at ritayan90@gmail.com

Location

Bengaluru

AI stack

claude code

codex

langchain

langfuse

langgraph

langsmith

Work experience

Technical Product Manager at Sigmoid, Oct 2024 - Present · 1 yr 11 mos, Bengaluru, Karnataka, India, Led 10+ AI-powered data products to GTM for enterprise CPG and pharma clients — owning each from discovery through launch. • Reckitt: shipped a data-product marketplace plus planogram and target/incentive/sales-hierarchy management tools. • Pfizer: built an AI-driven visit-pitch planner for medical representatives. • PepsiCo: delivered logistics tooling for truck-load optimisation. • Kenvue: built a data-product marketplace with data-sufficiency modules. • Prototype product UIs in Claude Design and ship v0 builds with Claude Code — delivering ~30% of the working product before engineering handoff for modularisation and scale, cutting requirement ambiguity and shortening feedback loops. • Design every AI-first product around enterprise-grade data security and privacy from the start — architecting data flows built to clear Architecture Review Board and Technical Review Board scrutiny, not retrofitted for compliance after the fact. • Translate enterprise client needs into AI product specs spanning RAG pipelines, agentic and multi-agent workflows, and model-routing strategy. • Own roadmap, prioritisation, and stakeholder alignment across data science, engineering, and client teams. • Apply a reliability-first lens from a decade in quality engineering — defining how non-deterministic AI features are tested, monitored, and gated in production.

QA Manager at rSTAR Technologies, Oct 2023 - Oct 2024 · 1 yr 1 mo, Bangalore Urban, Karnataka, India, Exelon- AI QA Lead - Lead a team of automation QAs to deliver high quality LLM based AI tools through high degree of automation coverage and test case generation. Innergy- QA Manager - Leading a team of manual and automation QAs to deliver high quality iOS/android apps and websites for both B2B and B2C verticals of the business

QA Manager at Tricog Health, Mar 2022 - Oct 2022 · 8 mos, Bengaluru, Karnataka, India

QA Lead at MediBuddy, Nov 2020 - Mar 2022 · 1 yr 5 mos, Bangalore Urban, Karnataka, India, Leading a team of 15+ members in manual testing, automation testing (ios,android and api) for both B2B and B2C wings of the product. Managing a suite of 7 products which includes iOS apps, Android apps, WebApps, integrated 3rd party mini apps for a market leading award winning healthcare app that's changing lifes across the country. Test Planning. Release management. Stake holder management. Have expertise in driving custom framework development for backend automation in python.

Assistant Quality Assurance Manager at DocsApp, Apr 2019 - Nov 2020 · 1 yr 8 mos, Greater Bengaluru Area, Docsapp merged with Medibuddy in 2020.

Education

Bachelor of Technology (B.Tech.), Electronics and Communications Engineering, West Bengal University of Technology, Kolkata

West bengal council of higher secondary education