How Sanity Context & MCP Supercharged AI Customer Support in Basegent & Bucket Space

Description and use

Integrating Sanity Context MCP into Basegent enables the AI assistant to provide accurate, condition‑aware responses, eliminating traditional RAG pitfalls.

Problem

Traditional vector search treats company documentation like a bag of text chunks, leading to outdated policy quoting, lack of conditional logic for multi‑dimensional entitlements, and hallucinations on generic product names.

Outcome

By modeling policies as structured documents in Sanity and querying them live via MCP, Basegent delivers accurate, context‑aware support, eliminates policy conflicts, and reduces hallucinations.

Workflow

• Integrate Sanity into Basegent via SanityContextSourceAdapter and BYOK inference engine
• Set up Sanity Knowledge Base & Context MCP with structured supportPolicy documents and conflict detection
• Deploy Basegent chat widget in Bucket Space, passing authenticated user tokens to query Sanity live

Evidence

Traditional vector search treats company documentation like a bag of text chunks.

Source

Support policies shouldn’t be plain blobs of text. In Sanity Studio, we modeled supportPolicy documents with explicit schema constraints.

Source

Built a first‑party SanityContextSourceAdapter directly into Basegent that talks to Sanity’s hosted MCP endpoint, paired with a multi‑provider Bring‑Your‑Own‑Key inference engine.

Source