Vendor-neutral signal pipeline architecture
Device adapter interface, stream normalization, signal quality gates, feature windows, and app-facing event output.
Platform
SlowSystems Neuro Platform
Product architecture for non-medical EEG processing boundaries, draft app-layer event models, signal quality gates, and separated technical environments.
Core platform areas
Draft app-layer cognitive-state event model
A minimized cognitive_state_update format for non-medical app/API workflows and reviewer discussion.
Defined feature extraction boundary
Signal quality status, relative band features, session window metadata, and reproducibility notes.
Separated environments
Public website, future app runtime, device adapter layer, analytics, and ingestion services remain distinct systems.
Output status
Initial prototype outputs
- Signal quality
- Relative band features
- Session stability
- Artifact flags
Experimental derived labels
- Focus-oriented
- Calm-oriented
- Attention-shift
Research questions
- Workload
- Mind-wandering
Technical stack separation
Signal pipeline stack
- Device adapter interface
- Stream normalization
- Timestamp handling
- Signal quality gates
- Feature windowing
- Draft event schema
- Test fixtures
- Reproducibility logs
Website stack
- React
- Vite multi-page build
- Cloudflare Pages
- GitHub
- Turnstile
- Resend
- Umami