Visualization Pipeline
HoloViews / Datashader / Dask rendering used by Gaius-era exploration UIs. Hub operators use signals-ui.
Stack
| Component | Role |
|---|---|
| HoloViews | Declarative data visualization — the agent composes views from the HoloViews gallery |
| Datashader | Server-side rasterization for large datasets at viewport resolution |
| Dask | Distributed parallel computation for on-demand view recomputation |
Agent-Mediated Flow
The gRPC engine directs this pipeline. When users issue commands in the WASM terminal, the agent:
- Interprets the instruction and selects a visualization strategy
- Directs Dask to perform distributed computation
- Datashader rasterizes the result at the current viewport resolution
- Updated HoloViews visualizations stream back to the web client
Capabilities
Resolution Autoscaling
Datashader renders data at the pixel resolution of the current viewport. On zoom:
- Dask recomputes the view in parallel over the narrowed data range
- Detail increases as the viewport narrows
- No client-side data transfer — only rasterized images transit the network
Multi-Persona Views
A single data product supports multiple persona conventions. The agent adjusts layout, metric selection, and visual conventions based on the active persona:
| Persona | Emphasis |
|---|---|
| Data Scientist | Statistical distributions, feature correlations |
| Domain Researcher | Domain-specific metrics, temporal patterns |
| Quantitative Analyst | Numerical precision, confidence intervals |
| Operational Analyst | System health, throughput, latency |
| Business Analyst | KPIs, trends, executive summaries |
MLOps Integration
The visualization pipeline supports:
- Feature extraction and correlation views
- Outlier detection and signal evaluation
- Signal correlation across temporal, spatial, and spectral dimensions
- Labeling workflows for different data modalities