Founding Product Engineer, Voice AI

Role Overview
We are hiring a Founding Product Engineer, Voice AI at TrainerTwin.
Chennai
₹15–25 LPA
3–7 years
Great coaches can only be in one room at a time. TrainerTwin changes that.
We turn a coach's knowledge, voice and teaching style into an AI twin learners can talk to anytime: role plays, live voice conversations and video, grounded in the real coach's material and approach.
We don't want to ship just another AI trainer. The hard, rewarding problem is a twin with high fidelity to the real coach. One that stays inside its boundaries, is stress-tested before learners meet it, and keeps getting better from every session and the coach's own review afterwards.
We are piloting with three international coaches and one in India, and we need an engineer to help build it end to end.
What It Looks Like
Ship features across backend and frontend, and watch how learners use them
Cut latency and fix turn-taking so the twin feels live
Build simulation-based evals that systematically test fidelity, boundaries and guardrails before launch
Feed learner sessions and coach reviews back into the twin (RLHF-style) so it improves with use
Who Is Right for This Role
You might fit if you've shipped real products, you're strong in Python, you're comfortable with async, streaming or realtime systems, and you measure before you optimize. Realtime video, gaming, telephony or high-concurrency backend work counts.
How You'll Work
We are a small, bootstrapped team working directly with the founders. Chennai-based.

Work arrangement
No