AI-powered industrial maintenance diagnostics, delivered through the browser and every chat app your technicians already use.
MIRA diagnoses equipment faults in conversation. Scan a QR code on a machine, tell MIRA what's wrong, and get an answer grounded in that specific asset's manuals, history, and the collective experience of every technician who came before. Built for the plant floor — mobile-first, voice-capable, CMMS-integrated.
- Live product: app.factorylm.com
- Marketing site: factorylm.com
- Product documentation: docs/product/
- Developer documentation: docs/developer/
- A technician walks up to a faulting machine.
- They scan a QR sticker on the equipment (or type in the asset tag, or upload a nameplate photo).
- MIRA opens a chat pre-loaded with the asset's vendor, model, service history, and recent faults.
- The technician describes the symptom — by voice or text.
- MIRA asks diagnostic questions, pulls the relevant page from the manual, and proposes a fix.
- When the fault is resolved, MIRA auto-generates the CMMS work order closeout. One tap to confirm.
Competitive differentiators: asset-scoped entry, industrial knowledge grounding (25K+ chunks across 100+ vendors), voice-first mobile UX, Atlas/MaintainX/Limble/Fiix CMMS integration.
- Docker + Docker Compose v2.20+
- Doppler CLI — all secrets via Doppler
- Ollama running on host at
localhost:11434 - macOS (Apple Silicon preferred) or Linux
- NeonDB account (free tier works for development)
Full setup walkthrough: docs/developer/local-setup.md
Proprietary — Copyright (c) 2026 Cranesync. All rights reserved. See LICENSE. Third-party bundled dependencies retain their original open-source licenses.
For licensing inquiries: mike@cranesync.com