Can AI agents automate Senzing?

Senzing entity resolution runs inside the customer's infrastructure as an embeddable SDK. This record is that engine. It is not an identity provider and not a CRM. The SDK page says it can be embedded in Python, Java, .NET, Go, or C++ and can also be deployed as a REST service or Docker container. A current public REST reference, with authentication and resource paths, was not opened, so the REST path is not scored. The v4 documentation hub publishes Python, Java, and C# SDK references and quickstarts for Linux, Docker, macOS, Windows, AWS, Azure, and air-gapped install. Subscription pricing on the opened page is $58,560 per year at 10 million data source records, under the standard end user license. The Senzing-owned skill documents a hosted MCP server at https://mcp.senzing.com/mcp, verified in the skill against server v1.37.13. Named tools cover mapping, local analysis scripts, docs search, SDK reference, code scaffolds, sample data, reporting guidance, and error codes. The same skill says the server works from pre-fetched documentation, never connects to a live Senzing instance, and never handles personal data. An evaluation license requested through the skill is a 10-day, 250,000-record trial. Connector catalogs were not opened. A product CLI was not established. Computer-use viability is unassessed. The hosted MCP scores 6 and is the best scored path because it is the documented agent interface. It does not resolve a live entity graph.

Best path today: mcp · Overall automatability: 6/10