From a messy real-world problem to something shipped and measured. I've written PRDs for voice and chat AI agents, designed RAG pipelines, and built a live games platform that turned a watching hall into ~150 players.
Real users, real data, my calls. Pick a story to play.
Six months taking voice and chat AI agents from discovery to release.
Sanyam combines a product manager's judgement with a genuine technical understanding of how modern AI systems are built, a combination that is rare at this stage of a career.
Everything on my resume, grouped the same way.
Literature and philosophy first, then product management with applied AI.
iHUB Divya Sampark, IIT Roorkee
AI capstone: Smart Split for PhonePeUniversity of Rajasthan
Jagadguru Ramanandacharya Rajasthan Sanskrit University
Board of Secondary Education, Rajasthan
Gujarat Secondary and Higher Secondary Education Board, Gandhinagar
How a literature student ended up shipping AI products.
I finished Class 10 with the Gujarat board in Gandhinagar and Class 12 with the Rajasthan board, studying English literature, Hindi literature, Sanskrit and Jain philosophy. Languages and ideas were my subjects long before products were.
A B.A. in English, Sanskrit and Jain philosophy, then an M.A. in English. Literature gave me the habit I use most today: noticing what a text, or a user, actually says rather than what I expected them to say.
IIT Roorkee's product management programme with applied AI showed me where that habit belongs. For my capstone I audited PhonePe's group-split flow, mined public reviews across five platforms, wrote the PRD for an AI bill-splitter, prototyped it in Figma and defended it before a panel.
As an AI Product Manager intern at Grand Slam Technologies, I owned requirements for voice and chat AI agents: build-ready PRDs, RAG pipelines with a retrieval-evaluation process, and a tested, version-controlled prompt library.
For a Jain NGO I turned a PowerPoint quiz into a KBC-style show, then built a games platform where the whole hall plays on their phones. When the show data showed right answers being marked wrong, and an hour of question-writing before every show, I shipped an AI question generator with a blind safety check.
I built it with Claude as my engineering partner. The problem, the decisions and the testing were mine.
I'm looking for an Associate Product Manager or AI Product Manager role where I can own a problem end to end: talk to users, write the spec, ship it, and check the data honestly.
Open to Associate Product Manager and AI Product Manager roles.
