About the Role :

Build AI-powered products end to end at Kenpath : PostgreSQL and APIs, React front ends, and the LLM, agentic layers behind them. You will work across government, social-sector and SaaS products where cost, latency and security matter from day one.

Responsibilities :

- Build backend services in TypeScript (NestJS) and Python (FastAPI) on PostgreSQL, and front ends in React and Next.js.

- Build agentic workflows with tool calling, structured outputs and MCP, using Pydantic AI, LangGraph or similar.

- Choose the right model per task across Claude, OpenAI, Gemini, Mistral and open-weight models, on quality, latency, cost and data residency.

- Evaluate AI features with golden datasets and regression tests, and optimise LLM cost and latency in production : caching, model routing, batching and per-feature cost tracking.

- Deploy on Cloud with CI/CD and infrastructure as code, and debug issues across database, API, model and deployment.

- Own modules from discovery to release, with clean, tested, documented code.

What You Bring :

- 3 to 6 years of software engineering, with at least 1 year of hands-on LLM work in production.

- Strong TypeScript, Node.js and Python, including async programming and API design.

- PostgreSQL, and Vector Database (collections, payload filters, hybrid search).

- Hands-on React, and working knowledge of AWS, Docker, Kubernetes and Git-based workflows.

- Experience building and measuring RAG systems, and LLM cost optimisation.

- Secure API practice : OAuth 2.0 / OpenID Connect, JWT and role-based access.

- An evidence-driven approach and clear written communication.