phase 14-15: docker build validation and helm deployment

This commit is contained in:
Celes Renata
2026-04-11 11:59:45 -07:00
parent 7394d241c9
commit ce10afa034
179 changed files with 32559 additions and 576 deletions
+15 -6
View File
@@ -3,9 +3,18 @@
Apache Superset dashboard configurations and starter datasets for Stonks Oracle.
## Starter Dashboards
- Symbol Overview — company profile, source health, recent documents
- Sentiment Heatmap — market-wide sentiment by sector and symbol
- Prediction Accuracy — predicted signals vs realized price moves
- Paper Trading PnL — paper trade performance and position tracking
- Model Quality — extraction success rates, latency, and confidence distributions
- Source Coverage — ingestion throughput, source failures, and coverage gaps
See `starter/` for dashboard definitions covering:
- Symbol Overview — company profiles, source health, recent documents, and market snapshots
- Sentiment Heatmap — market-wide sentiment by sector and symbol, catalyst analysis
- Prediction Accuracy — predicted signals vs realized price moves, confidence calibration
- Paper Trading PnL — cumulative PnL, position snapshots, order history, and scorecards
## Operational Dashboards
See `operational/` for dashboard definitions covering:
- Ingestion Throughput — documents/hour by source type, success/failure rates, stale sources
- Model Extraction Quality — success rates, latency percentiles, validation failures, confidence distributions
- Source Coverage & Gaps — per-symbol source type matrix, missing sources, failure heatmap
Starter dashboards are powered by the Trino `lakehouse` catalog over MinIO-backed analytical tables.
Operational dashboards query the Query API `/api/ops/*` endpoints.
All dashboards can be imported into Superset via the UI or CLI.