How vijay-rodge/bis_manak works
This is a Streamlit assistant for finding Indian Standards and explaining related BIS certification and consumer guidance. Its primary workflow takes a product or standard query, optionally scrapes and stores live BIS results, ranks matching standards, and produces a technical extraction followed by a conversational answer. Separate UI branches support manual scraping, database exploration, and reference guides. The README documents these capabilities; some unsampled UI wiring and internal LLM-stage details are represented only where documented.
Main entry points: app.py Streamlit App · search_loop.py Search Orchestrator · scraper.py Portal Scraper · tab_scraper.py Scraper Tab
Components 17
User Experience
- Streamlit App
app.py - Assistant Chat
ui/tab_chat.py - Scraper Tab
ui/tab_scraper.py - Database Explorer
ui/tab_database.py - BIS Guide
ui/tab_guide.py
Standards Discovery
- Search Orchestrator
search_loop.py - Portal Scraper
scraper.py - Standards Matcher
retriever.py - Technical Extractor
llm_pipeline.py - BIS Advisor
llm_pipeline.py
BIS Knowledge
- BIS Domain Guides
bis_knowledge.py
Standards Data
- Standards Store
db.py - Standards Dataset
standards_data.csv - Database Seeder
seed_database.py
Other components
- User
- BIS Standards Portal
- LLM Providers
Connections 22
- User → Streamlit App: opens
README.md - Streamlit App → Assistant Chat: shows
README.md - Streamlit App → Scraper Tab: shows
README.md - Streamlit App → Database Explorer: shows
README.md - Streamlit App → BIS Guide: shows
README.md - Assistant Chat → Search Orchestrator: runs query
ui/tab_chat.py - Search Orchestrator → Portal Scraper: scrapes
search_loop.py - Portal Scraper → BIS Standards Portal: fetches results
scraper.py - Portal Scraper → Standards Dataset: saves records
scraper.py - Search Orchestrator → Standards Store: saves records
search_loop.py - Search Orchestrator → Standards Matcher: ranks candidates
search_loop.py - Standards Matcher → Standards Store: loads records
retriever.py - Search Orchestrator → Technical Extractor: runs extraction
README.md - Technical Extractor → BIS Advisor: supplies findings
README.md - BIS Advisor → BIS Domain Guides: uses scheme rules
llm_pipeline.py - Technical Extractor → LLM Providers: calls model
llm_pipeline.py - BIS Advisor → LLM Providers: calls model
llm_pipeline.py - Scraper Tab → Portal Scraper: scrapes on demand
ui/tab_scraper.py - Scraper Tab → Standards Dataset: saves records
ui/tab_scraper.py - Database Explorer → Standards Dataset: explores data
README.md - BIS Guide → BIS Domain Guides: presents guides
README.md - Database Seeder → Portal Scraper: uses scraper
seed_database.py