We're building an AI platform for credit unions. A variety of tools (third party integrations + RAG) backed by multiple LLM options, including OLMo, running in our own AWS infrastructure. We're a small team (I'm CTO and our first technical hire) so you'd have real ownership over the AI stack and architecture.
You'd work on retrieval over heterogeneous corpora with citations, agent orchestration and tool calling, model serving infra, and third party API integration. There's a lot of room to experiment with hierarchical retrieval, heavier upfront document processing, and custom agent tooling. Over time we're going to expand into fine tuning, more direct compute management, and some interpretability work.
Stack: Python/FastAPI, React, PostgreSQL/pgvector, AWS/OpenTofu. Looking for someone with a few years of experience building production systems.