The Challenge
Fleet data lived inside the application, so routine questions — how many vehicles are currently in service, which vehicle is on which trip — meant opening the system and reading a report every time.
Case study
A Frappe fleet system with a custom MCP layer, so vehicle and trip status can be asked for in plain language inside Claude or ChatGPT.
Fleet data lived inside the application, so routine questions — how many vehicles are currently in service, which vehicle is on which trip — meant opening the system and reading a report every time.
A Frappe application holding vehicles, trips and service state, paired with a custom MCP server that exposes that data as scoped tools. An assistant like Claude or ChatGPT can then answer fleet questions directly in chat, reading live records rather than a stale export.
Frappe and Python for the fleet data model, with a Model Context Protocol server bridging it to MCP-capable AI clients.
More work
Tell us what you're trying to automate, integrate or build. We'll tell you what it would take.