Product Manager - Agentic Systems

Location : Gurugram

Experience : 4 - 7 years

About TrueFan AI :

TrueFan AI is an enterprise platform for personalised AI video. Enterprises use it to turn one recording into thousands of video variations, for marketing, sales, customer engagement and internal communications. In 2025 we generated 20M+ videos for 100+ enterprises, more than 15 of them with market caps above $10B. We raised a $10M Series A in June 2026.

The Role :

This is a new bet for us, and there is no playbook. You will get a hard problem, a small team of engineers, and the freedom to work out what to build. We are building agentic systems that do work which used to take teams of people : writing, storyboarding, generating, editing and checking video at enterprise scale. You will decide what these systems do, how they are designed, and what becomes a product. Where you will start : content at scale. Enterprises want thousands of unique videos generated from their own data, and today each one needs a custom agentic pipeline. You will design and ship these pipelines with four engineers, find what repeats across them, and build the product that removes most of the custom work. Where it goes from there is largely up to you.

What You will Do :

- Design agentic systems : the steps, models, tools, QC loops and how failures are handled

- Scope client use cases with our delivery team, and produce cost, feasibility and timeline estimates you can stand behind

- Set the team's direction and make the product calls

- Prototype agents and prompts yourself before engineering builds them

- Define how quality is measured, and build evals with the team

- Turn patterns across client builds into reusable components and products

Must-Have :

- 4 - 7 years of experience, including time as an engineer or in a hands-on technical role

- Have built agentic or multi-step LLM systems yourself, not just written specs for them

- Able to read code, review system designs, and reason about cost, latency and failure rates

- Took something from zero to shipped with little direction

- Comfortable explaining technical tradeoffs to enterprise clients

- Clear written communication : specs and decision docs

Good to Have :

- Experience with generative image, video or voice models

- Experience building evals or LLM-as-judge systems

- Experience turning repeated custom builds into a product

You Are :

- Someone who works best with a problem, not a roadmap

- Comfortable making decisions with incomplete information, and changing them when the data says so

- A builder first : you would rather prototype an idea than write a deck about it

- Honest about what can be built, especially in front of a client

- Excited by how fast this field moves rather than tired by it.