Full-Stack AI Engineer (MVP Development, Voice & RAG Systems)

Echotree is a legacy storytelling platform that creates interactive AI personas called Echoes, built from a person’s voice, memories, and life story. We facilitate long-form interviews with clients where we gather ~80,000 words of data to create an AI reflection of the individual that family members and friends can interact with. For example, a grandchild can ask their grandparent’s Echo about their wedding day and hear all the details back in their same voice.

We already have a working proof of concept (text-based Echo chat using Gemini + Supabase + Vercel). Now we’re looking for a strong developer to help us bring the next layer to life.

This role will begin on a project basis, building out the next critical features of Echotree. If it’s a great fit, there’s a clear path to transition into a long-term role with potential equity as a founding engineer.

You’ll be responsible for:

• Integrating voice playback into the chat UI (with fallback, caching, and per-Echo account mapping).
• Adding a per-Echo voice usage meter and quota enforcement system (hard cap, usage tracking, alerts).
• Implementing basic monitoring/logging (Supabase logs, error handling, Slack alerts).
• Extending our existing Supabase/Vercel/Gemini backend.
• Polishing our UI and smoothing out bugs/inconsistencies.
• Laying groundwork for subscription automation (free trial logic, tier enforcement, Stripe integration).
• Optimizing our RAG pipeline (embedding transcripts, semantic retrieval, reducing Gemini API costs).

This is a hands-on coding role with a lot of ownership. We’re looking for someone who can move quickly, is interested in the larger mission of the project, and can grow alongside Echotree as we scale.