🕉
Minimum Viable Product (MVP)
The MVP is a conversational AI chat interface where users ask questions about the Vedas and receive contextual, referenced answers powered by web search and Sarvam AI.
Try it live: veda-guru-ai-ui.vercel.app
Scope
| Area | Included in MVP | Future |
|---|---|---|
| Chat | Text-based Q&A with AI | Voice, guided tutorials |
| Search | Web search (DuckDuckGo + Google fallback) | Curated Vedic knowledge base |
| LLM | Sarvam AI (sarvam-105b, 128K context) |
Fine-tuned Vedic model |
| Languages | English | 10+ Indian languages |
| References | Source URLs from web search | Clickable verse references |
| Security | Rate limiting, CORS lock, API key auth | User auth, per-session tokens |
| Learning | Free-form Q&A | Structured courses, progress tracking |
How it works
- User asks a question (e.g. “What does the Rig Veda say about creation?”)
- Backend searches the web via DuckDuckGo (falls back to Google)
- Search results are fed as context to Sarvam AI (
sarvam-105b) - Sarvam AI synthesizes an answer with source citations
- Response is rendered with markdown formatting in the chat UI
User flow
Ask a question
│
▼
Web search (DDG → Google fallback)
│
▼
Context + question sent to Sarvam AI
│
▼
AI synthesizes answer with source citations
│
▼
Response rendered with markdown in chat UI
Tech stack (MVP)
| Component | Choice |
|---|---|
| LLM | Sarvam AI (sarvam-105b, 128K context) |
| Backend | Python FastAPI (Render — free tier) |
| Frontend | Vanilla HTML/CSS/JS (Vercel — free tier) |
| Web search | DuckDuckGo + Google fallback (no API key needed) |
| Rate limiting | slowapi (30 req/min per IP) |
| Markdown render | marked.js |
Success metrics
- Accuracy: Responses are factually grounded in search results (>90% precision)
- Relevance: User finds the answer helpful (>80% satisfaction)
- Latency: First response within 5s (or up to 60s on cold start)