Conversational commerce
Azure OpenAI
Retrieval pipelines
RBAC & governance
Role & scope
I led the project from discovery through to a production-ready architecture. The focus was on connecting product catalogues, CRM and the pricing engine so the chatbot delivers context-rich answers and takes work off sales teams' hands.
- Discovery workshops to sharpen use cases, data structures and risks.
- RAG pipeline with chunking, enrichment and ranking based on up-to-date product data.
- Azure OpenAI with guardrails, an eval harness and observability to safeguard quality.
- RBAC and an audit trail so commercial and legal teams can collaborate compliantly.
Impact
- Quote turnaround time cut from days to hours through automatic summarisation and task routing.
- Higher conversion through consistent follow-up and the suggestion of cross/upsell alternatives.
- A monitoring dashboard provides real-time insight into quality, usage and future optimisations.