Midday is joining Ramp
We started Midday to build something we wanted for ourselves and it grew into something much bigger than we expected.
And here is the story behind it 🧵
This is how I structure our multi-agent AI system:
• 10 agents (triage + 9 specialists)
• 43 tools (grouped by domain)
• 12 artifacts (visual canvases)
Each agent gets only the tools it needs. Clean and maintainable.
The best possible OCR solution for invoice extraction is a combination:
- Mistral OCR as primary
- Gemini 2.5 Flash
- Fallback to text extract
- Combine results