Hiring: Senior Full-Stack Engineer (AI) – Evatt AI
Remote, full-time contractor (40 hrs/week) → possible conversion to full-time + long-term option to relocate to Australia
Must be within ±3h of GMT+8 (India, Singapore, China, Malaysia, WA)
About us
Evatt AI is building AI tools for lawyers. Current stack is Next.js + React + TypeScript on the app side, and Python/FastAPI + vector search + LLM/RAG on the AI side. Next phase is to build a legal casebase/search product similar to JADE.io / AustLII (natural-language search over case law and legislation). You will work directly with the founder and own delivery.
What you’ll do
-
Own the codebase (Next.js, FastAPI, Docker microservices)
-
Build the legal casebase (RAG + vector DB such as Pinecone/Qdrant)
-
Improve AI streaming/retrieval
-
Refactor UI into modular React components
-
Ship, test, deploy, keep staging/prod stable
Tech we need
-
Next.js 15, React 19, Tailwind, MUI
-
Node.js, TypeScript, Drizzle ORM, Zustand
-
Python 3.11+, FastAPI, Pydantic
-
Postgres/MySQL
-
Pinecone (Qdrant/Milvus a plus)
-
LLM APIs: OpenRouter / OpenAI / Gemini / Claude
-
Docker, Railway, Stripe, Google OAuth, SendGrid
Nice to have: LangChain/LlamaIndex, Elasticsearch/Weaviate, CI/CD (GitHub Actions), performance tuning.
Interview project
Small prototype: upload 10–20 legal cases → embed to vector DB → natural-language query (e.g. “breach of contract in retail”) → return ranked snippets. Clear architecture + clean code + good retrieval = pass.
Apply
Email ashley@evatt.ai
Subject: Evatt AI – Full-Stack AI Engineer Application
Include: short intro, GitHub/portfolio, and (optional but preferred) 3–8 lines on how you’d build the JADE.io/AustLII-style search.