Connect Claude, GPT-4, or any MCP-compatible AI directly to your live Diba platform — headcount, access events, and equipment sensors — queryable in plain language.
MCP is an open standard that lets any AI assistant securely call external tools and read live data. One Diba MCP server makes your entire platform — access events, headcount, IoT sensors — queryable by any compatible AI.
Open standard · Works with any MCP-compatible AI client
Six live tools exposed via MCP. Any AI can call them — no custom coding on your end.
Instant worker counts per gate, zone, or across all sites — updated in real time as workers scan in and out.
Retrieve filtered access events by worker, gate, time window, or decision — granted, denied, or flagged.
Check if a specific worker is currently on site, when they arrived, which gate they used, and their induction status.
Read live sensor data from any asset on the Edge Gateway — temperature, pressure, run state, and active alarms.
Generate an evacuation-ready muster list — live entry times, zones, and contact details — on demand, per site.
Query site-wide induction rates and surface any workers with expired credentials before they reach the gate.
Your AI calls Diba's MCP tools automatically — no query language, no dashboards, no waiting.
No new hardware. Works alongside your existing Diba platform.
Diba provisions an MCP server tied to your platform API. It runs in our cloud — nothing to install, nothing to configure on your end.
Add the Diba server to Claude Desktop, Cursor, or any MCP-compatible client. One config block — takes about 60 seconds.
Your AI has structured, live access to every part of your platform. Ask in plain English — responses arrive with real data in under a second.
Diba MCP is in private early access. Tell us about your sites and AI stack — we'll onboard you personally within a week.