How medintelx-healthcare-services-pvt-ltd/dataminingdiscovery works
This is a Streamlit healthcare-claims audit training and questionnaire application. Users select an audit concept, configure or load questions, review claim examples, answer a quiz, and receive a score; the app also supports synthetic claim generation and schema exploration. The graph includes the core data, persistence, and AI-generation layers. The README is empty, and app.py excerpts are incomplete, so some likely UI-to-engine connections remain unverified rather than being presented as confirmed calls.
Main entry points: app.py Streamlit application · schema_manager.py Schema manager · llm_synthetic_generator.py AI generation · schema_reader.py Schema reader
Components 15
Audit experience
- Streamlit application
app.py - UI components
modules/ui_components.py - Concept rules
config - Questionnaire builder
modules/questionnaire_builder.py - Quiz evaluation
modules/evaluation_engine.py
Scenario generation
- Rule-based scenarios
modules/synthetic_data_generator.py - AI generation
modules/llm_synthetic_generator.py
Schema and data
- Schema reader
modules/schema_reader.py - Schema manager
modules/schema_manager.py - ClickHouse client
modules/clickhouse_client.py
Local persistence
- Questionnaire store
config/questionnaires.db - Schema cache
config/schema_cache.db
Other components
- Auditor
- ClickHouse
- Azure OpenAI
Connections 16
- Auditor → Streamlit application: uses
app.py - Streamlit application → UI components: renders with
app.py - Streamlit application → Concept rules: loads rules
app.py - Streamlit application → Questionnaire builder: loads saved questions
app.py - Streamlit application → Quiz evaluation: evaluates answers
app.py - Streamlit application → Rule-based scenarios: generates scenarios
app.py - Streamlit application → AI generation: requests AI generation
app.py - Streamlit application → Schema reader: explores schema
app.py - Questionnaire builder → Questionnaire store: reads and writes
modules/questionnaire_builder.py - Schema reader → Schema manager: gets table schema
modules/schema_reader.py - Schema reader → ClickHouse client: queries samples
modules/schema_reader.py - Schema manager → Schema cache: reads and writes
modules/schema_manager.py - Schema manager → ClickHouse client: fetches columns
modules/schema_manager.py - ClickHouse client → ClickHouse: queries
modules/clickhouse_client.py - AI generation → Schema manager: gets friendly names
modules/llm_synthetic_generator.py - AI generation → Azure OpenAI: requests completions
modules/llm_synthetic_generator.py