Turn enterprise APIs into governed AI tools
AI API Mapper sits between the business APIs you already run and the agents that want to call them. Import OpenAPI, SOAP/WSDL, and OData v4 contracts — Microsoft Graph included — map and enrich them into business-friendly tools, then publish governed toolsets over MCP and A2A, with SDKs for .NET, TypeScript, Python, and Java.
From onboarding to audited execution
AI clients can't safely call the APIs you already built. Downstream APIs expose technical operations, inconsistent payloads, and authentication schemes that were never designed for AI tool use — while AI clients need clear names, typed inputs, predictable outputs, and a controlled execution boundary.
AI API Mapper closes that gap end to end: import, map, enrich, publish, govern, invoke, audit. The result is controlled enterprise AI tool exposure — not simple API proxying.
Key capabilities
Nine capability areas — from contract onboarding to SDK integration — designed to give enterprise teams full control over how AI agents interact with the systems they already run.
API Onboarding
Import OpenAPI, SOAP/WSDL, and OData v4 contracts, Microsoft Graph included. Preview every candidate operation before anything is persisted, and keep revisions side by side as API version families with breaking-change comparison.
Mapping Lifecycle
Draft, propose, review, approve, publish, deprecate, revert, and restore — every step versioned. Exposed field names form the contract your agents see; the source API's own names stay an internal detail.
AI Enrichment
LLM-assisted names, descriptions, and field suggestions generated from your onboarded contracts. Large payloads are split across passes rather than truncated, shared types are named once, and runs are deterministic and resumable.
Governed Toolsets
Group mappings into toolsets and assign them to client applications, each with its own prompt composition and schema discovery mode — so one tenant can publish different tool catalogs to different AI consumers.
MCP & A2A
Serve the same governed catalog over MCP 2025-11-25 with Streamable HTTP, and over A2A 1.0 as a concentrator or a hosted agent. One catalog, reshaped — never recomputed.
Credential Binding
Downstream authentication from passthrough bearer and API keys to SOAP UsernameToken, mTLS client certificates, and token exchange. Secrets resolve from managed platform storage or Azure Key Vault, inside infrastructure you control.
Runtime Governance
Default-deny tool authorization: when no active policy matches, the call does not run. Tenant-aware policies match on issuer, client ID, roles, scopes, groups, and custom claims.
Audit & Traceability
Inspect any call end to end by correlation ID with per-step timing, with opt-in payload capture that is sanitized and truncated automatically. Deliver audit records to external HTTPS, TLS syslog, SQL Server, and PostgreSQL sinks.
SDKs & Adapters
Client libraries for .NET, TypeScript, Python, and Java, with adapters for Microsoft.Extensions.AI, Semantic Kernel, Azure AI Foundry, OpenAI, LangChain, LangGraph, and Spring AI. Get the system prompt, get the tools, invoke a tool.
Now includes ESRA
ESRA — Enterprise Semantic Retrieval Agent lets an AI client ask a question in plain language and receive governed evidence from your own enterprise databases. It is published through the same Runtime as an ordinary MCP tool, so it composes into your existing tool catalog additively — and it is included on every plan at no additional cost.
Discover ESRAReady to expose APIs to agents safely?
Get a technical walkthrough with our team — or explore the platform documentation at your own pace.