An end-to-end agentic system that ingests SOP documents, generates audit workflows, binds rules, and executes claims decisions — built for Wipro & UHC Optum.
A 122-agent LangGraph BFS pipeline crawls HTML, DOCX, XLSX, and PDF documents — enriches with LLMs — and writes to a multi-store knowledge graph.
Auditors assemble claim-audit workflows on a visual canvas — no code, no SQL, no engineering dependency.
property_schema drives the inspector form — adding a field requires no frontend deployment.PUT /workflow/:id/graph/ persists the entire nested graph transactionally to Postgres.Each workflow shape exposes an "attachable" picker — auditors select exactly which SOP rules and API agents apply at that decision point.
GET /workflow/:id/attachable/tool_calls JSONB — invoked automatically during execution.A 7-node LangGraph pipeline walks the bound workflow, evaluates each rule against the claim, calls external APIs, and produces a fully cited decision.
The AI proposes. The auditor decides. Every escalated or borderline claim lands in the HITL Review Dashboard for human override.
Toystack AI designed and built a unified Model Context Protocol for all tool calls in the execution engine — standardising auth, logging, schema validation, and retry.
Tool call responses are huge JSON blobs. Stuffing them into LLM context windows is expensive and lossy. Graph MCP stores each response as a Neo4j subgraph — the LLM queries the graph instead of receiving the full payload.
Full API response (often 50–200 KB) injected directly into the LLM prompt. Context window fills up. Relevant fields are buried. Costs spike per call.
Response is stored as a typed graph. The LLM receives only a schema summary and runs Cypher queries to fetch exactly the nodes it needs. Context stays small.