AI Engineer London | Joshua Fields
London-based AI engineer building production systems where language models call tools, manage state, use APIs, recover from failures, and support real user workflows.
Best fit
- AI products that need full-stack engineering, not isolated prompt demos.
- Voice/video AI assistants using LiveKit, STT/TTS, tool-calling, and backend orchestration.
- LLM workflows that turn transcripts, documents, or operational knowledge into structured outputs.
Proof
- Built autonomous tool-driven agents at Venaglass across internal platforms and external APIs.
- Built Emma AI workflows that transformed transcripts into reports, notes, summaries, and email drafts.
- Built Needle with real-time voice onboarding, transcript-driven enrichment, self-hosted LLMs, and explainable matching.
- Built RAG and vector-search projects using FastAPI, LangChain, FAISS, Supabase, and pgvector.
Stack: OpenAI, Anthropic, Hugging Face, LangChain, LangGraph, LiveKit, FastAPI, Python, React, Next.js, Supabase, pgvector
Download AI engineer CV | Email Joshua | AI projects
Joshua Fields — full portfolio
