Spaces:
Sleeping
Sleeping
A newer version of the Streamlit SDK is available:
1.52.1
metadata
title: SAP Chatbot
emoji: π€
colorFrom: blue
colorTo: purple
sdk: streamlit
sdk_version: 1.28.0
app_file: app.py
pinned: false
π§© SAP Intelligent Assistant
A free, open-source RAG (Retrieval-Augmented Generation) system for answering SAP-related questions using cloud LLMs and vector databases.
β¨ Key Features
- β 100% Free & Open Source
- β Multi-source SAP data (Community, GitHub, StackOverflow, Dev.to, Medium)
- β Production-ready: Supabase + pgvector vector database
- β HuggingFace Inference API for fast responses
- β Automatic data ingestion via GitHub Actions
- β Beautiful Streamlit UI
- β Multi-user cloud hosting
- β Conversation history with source attribution
π How It Works
1. Data Collection β 2. Embeddings β 3. Vector Search β 4. Answer Generation
(SAP sources) (sentence- (Supabase (HF Inference
transformers) pgvector) API)
Supported Topics:
- SAP Basis Administration
- SAP ABAP Development
- SAP HANA
- SAP Fiori & UI5
- SAP Security & Authorization
- SAP BTP (Business Technology Platform)
- SAP Integration Suite
- SAP Performance Tuning
- And more!
π§ Setup
1. Local Development (with Ollama)
# Clone repo
git clone https://github.com/Akshay-S-PY/sap-chatboot
cd sap-chatboot
# Create virtual environment
python -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install -r requirements.txt
# Build dataset
python tools/build_dataset.py
# Run locally
streamlit run app.py
2. Production (Supabase + HF Spaces)
See SUPABASE_SETUP.md for step-by-step cloud deployment.
π Architecture
GitHub Repository (sap-chatboot)
β
GitHub Actions Workflows:
1. build_dataset.yml β Dataset + Upload to HF Hub
2. ingest.yml β Ingest to Supabase
3. deploy_spaces.yml β Deploy to HF Spaces
β
Supabase Database (pgvector + RLS)
β
Streamlit App (HF Spaces)
β
User Query β Vector Search β LLM Response + Sources
π Tech Stack
| Component | Technology | Cost |
|---|---|---|
| Vector Database | Supabase (pgvector) | Free |
| Embeddings | sentence-transformers | Free |
| LLM API | HuggingFace Inference | Free |
| App Hosting | HF Spaces | Free |
| Data Pipeline | GitHub Actions | Free |
π‘ Use Cases
- Quick SAP Questions: Get instant answers about SAP config, ABAP, Basis
- Learning: Understand SAP concepts with cited sources
- Team Knowledge Base: Share with your entire team
- Integration: Use programmatically via Python API
π Resources
- π GitHub Repository
- π Supabase
- π€ HuggingFace
- π¬ SAP Community
β οΈ Important Notes
- First run builds dataset (~5-10 min)
- Works 100% offline with Ollama
- All data sources are publicly available and respectfully scraped
- No personal data is stored
Made with β€οΈ for the SAP Community
Have questions? Check the documentation or create an issue.