We’re building a 24/7 AI-powered financial co-pilot that users rely on for life. Deep expertise in personal finance, financial planning, and financial analysis, backed by the data infrastructure to make it trustworthy at scale. You’ll own both the intelligence layer and the database backbone.
Compensation: Highly paid work with the opportunity to be part of “the next unicorn” in stealth mode. Salary range: $50-$70/hour plus equity
What you’ll work on
- PostgreSQL & AlloyDB — you’ll own schema design, indexing, partitioning, query optimization, and the migration
- BigQuery data layer — building out analytics pipelines on GCP
- AI & LLM integration — RAG, Vertex AI Agent Garden, LangGraph/LangChain orchestration, conversation memory, Google Gemini in production
- Full-stack features — FastAPI routes → Postgres/Alembic migrations → Next.js UI
- Evals & reliability — LangSmith regression testing for financial math, Monte Carlo engine, SOC 2 alignment
- Multi-tenant platform & integrations — org/permissions system, 3rd-party integration, CRM systems, etc.
Stack Python/FastAPI · Next.js/TypeScript · PostgreSQL & AlloyDB · LangGraph/LangChain · Google Gemini/Vertex AI · GCP (Cloud Run · Cloud SQL · Cloud Build · BigQuery) · Auth0 · LangSmith
You’re a fit if
- Expert PostgreSQL — indexing, partitioning, performance tuning, migrations at scale
- You’ve shipped production financial or data-sensitive software and understand the stakes
- Hands-on with GCP data stack (AlloyDB, BigQuery, Cloud SQL)
- You’ve used LLMs in production — you know the difference between “it seems to work” and “the eval passed”
- Comfortable owning a FastAPI service and the Next.js component
- Curious about making personal finance accessible to everyone as someone who thinks people deserve better tools
- Must have US work authorization; must be able to pass a background check
Not a fit if
- You need design specs and fully-scoped tickets before starting
- You’re a pure specialist (backend-only or ML-only)
- You’d rather ship features than validate them
