At AdvanceMed we walked through the Deep Breath platform architecture: from BreathBox collection at the bedside, through real-time and archive storage, to application layers for research, education, and clinical decision support.
The talk emphasised hospital-controlled deployment, time synchronisation across devices, and a vendor-agnostic design that fits both academic centres and smaller hospitals.
Architecture layers
- Collection — multimodal streams from ICU devices
- Storage — real-time + archive database
- Processing — Deep Breath AI, hospital AI, third-party models
- Interfaces — decision support, research, and education UIs
