AIChatbotProductFamily
RoleFull-Stack Engineer
YearEarly 2026 — Now
CategoryAI / LLM
StatusShipped
Full-stack engineer on a production AI chatbot product family — a suite of conversational and lead-generation widgets that businesses embed on their own sites, powered by retrieval-augmented LLMs.
I work as a full-stack engineer on an AI chatbot product family used by real businesses in production, building it alongside an external dev team. It's a suite — a conversational chatbot, a lead-capture widget and a richer 'pro' widget — all sharing one AI core and embeddable on any website.
On the backend I build the retrieval pipeline: crawling and ingesting customer content, generating embeddings, storing them in a Postgres + pgvector index, and feeding the right context into the LLM so answers stay grounded in each customer's own material. Background processing runs through Celery; everything is containerised with Docker and shipped to cloud infrastructure.
On the frontend I build the embeddable widgets and the admin dashboard where customers configure their bots, tune prompts and review conversations. The work spans the full stack — data model to deploy pipeline to the pixel a visitor clicks.
Beyond code I work across architecture, performance and reliability of the AI features, and help coordinate the external dev team — turning fuzzy product ideas into shippable systems.