Samsara launches MCP to link fleet data with over 40 AI tools
On Sept. 22, Samsara launched its Model Context Protocol (MCP), offering more than 40 read-only tools that let authorized users query live fleet data via AI tools such as ChatGPT and Claude without custom integration. Polyak Trucking used Claude via MCP to identify a driver spending over 110 minutes per shift in the yard, recovering $53,000 in labor costs.
MCP provides read‑only access to live data on vehicles, drivers, safety events, hours of service and other operational metrics. Users can retrieve this information through AI applications such as ChatGPT, Claude or Microsoft Copilot without needing a custom integration. Samsara said MCP is free to use now, with future pricing based on monthly usage.
“Our customers are building AI into more of their business, and decisions in areas like finance and HR need to reflect what’s happening in the field. Samsara provides that operational understanding. With MCP, customers can bring their own Samsara data into the systems and AI tools those teams already use, connecting the work of running their operations with the work of running their business,” O’Driscoll said.
“With MCP connections already in place across our other major business systems, we can now bring Samsara’s operational context into those same workflows. That lets our operators focus on solving business problems instead of maintaining yet another custom integration,” Ken Apple, CTO at Mariner Logistics said.
Before MCP, Samsara offered Agent Studio for building agents inside its platform and maintained over 400 integrations, APIs and developer tools. MCP is built on the same live operational data layer that powers Samsara Assistant.
What we know
Samsara announced on Sept. 22 that its new Samsara Model Context Protocol (MCP) allows authorized users to query live information from their Samsara accounts through third-party AI applications without building a custom integration.
More than 40 read-only tools are available through MCP at launch.
She asked the AI tool to identify drivers spending more than 30 minutes on duty without driving while at the company’s terminal.
Polyak said the company recovered $53,000 in labor costs after identifying the issue.
She also described using Claude to compare toll costs between drivers running the same lanes, which uncovered a $50-per-load difference.