How ankitvarma11/cews works
CEWS is a local competitive-intelligence system for life-sciences organizations: it gathers public evidence, resolves it into organizations and topics, computes explainable scores and forecasts, and presents results for human review. The documented analyst workflow runs from source collection through normalization, competitor discovery, feature generation, scoring, and forecasting. Supporting capabilities include deterministic insights, a dashboard, API, exports, scheduled runs, and optional AI-assisted chat. Some subsystems are present in the tree but were not sampled; their links are omitted or marked unverified rather than inferred.
Main entry points: cli.py Command workflow · models.py Analytical database · pipeline.py Feature generation · app.py Streamlit dashboard
Components 20
Evidence collection
- Source ingestion
src/cews/ingestion/orchestrator.py - Public data adapters
src/cews/ingestion/adapters - Analytical database
src/cews/database/models.py
Entity processing
- Record normalization
src/cews/normalization/pipeline.py - Human review actions
src/cews/normalization/review_actions.py - Competitor discovery
src/cews/discovery/competitor_discovery.py
Analytics
- Feature generation
src/cews/features/pipeline.py - Explainable scoring
src/cews/scoring/pipeline.py - Forecasts and anomalies
src/cews/forecasting/pipeline.py
Insights and delivery
- Deterministic insights
src/cews/insights/rule_engine.py - Dashboard data queries
src/cews/dashboard/queries.py - CSV and Power BI exports
src/cews/exports - Scheduled jobs
src/cews/scheduler
User interfaces
- Command workflow
src/cews/cli.py - Streamlit dashboard
dashboards/streamlit/app.py - Review panel
dashboards/streamlit/components/review_panel.py - HTTP API
src/cews/api/app.py - Analyst chat
dashboards/streamlit/components/chat_widget.py - Chat agent runtime
src/cews/agents/session.py
Other components
- Analyst
Connections 20
- Analyst → Command workflow: runs commands
README.md - Command workflow → Source ingestion: invokes collection
src/cews/cli.py - Source ingestion → Public data adapters: dispatches sources
- Command workflow → Record normalization: runs normalization
README.md - Command workflow → Competitor discovery: runs discovery
README.md - Command workflow → Feature generation: runs features
README.md - Feature generation → Analytical database: stores features
src/cews/features/pipeline.py - Command workflow → Explainable scoring: runs scoring
README.md - Feature generation → Explainable scoring: supplies inputs
src/cews/scoring/pipeline.py - Command workflow → Forecasts and anomalies: runs forecasts
README.md - Forecasts and anomalies → Analytical database: stores forecasts
src/cews/forecasting/pipeline.py - Deterministic insights → Analytical database: reads stored scores
src/cews/insights/rule_engine.py - Streamlit dashboard → Dashboard data queries: requests data
dashboards/streamlit/app.py - Dashboard data queries → Analytical database: reads records
src/cews/dashboard/queries.py - Review panel → Human review actions: applies decisions
dashboards/streamlit/components/review_panel.py - Command workflow → HTTP API: serves API
- Streamlit dashboard → Analyst chat: offers chat
dashboards/streamlit/app.py - Analyst chat → Chat agent runtime: runs agent
dashboards/streamlit/components/chat_widget.py - Scheduled jobs → Command workflow: runs workflow
src/cews/cli.py - CSV and Power BI exports → Analytical database: exports results