RagaBot
AI-Powered Indian Classical Music Platform
What We Built
RagaBot is an AI-powered research and knowledge management platform for Indian Classical Music. It integrates OpenAI, Gemini, and Perplexity with web scraping and human verification workflows to build the most comprehensive repository of verified musical metadata - enabling structured insights on artists, raags, and taals by blending traditional knowledge with cutting-edge AI.
The Problem We Solved
Indian Classical Music has centuries of rich history, but its knowledge base remains fragmented across books, archives, and oral traditions. Researchers and institutions like Saptak Archives struggled to digitize, organize, and verify vast amounts of musical metadata - from artist biographies to intricate raag structures.
Key Pain Points
- Scattered data across thousands of unstructured sources
- No unified platform to search and verify musical metadata
- Manual research taking weeks per artist or raag entry
- Inconsistencies and inaccuracies in available data
- Lack of collaboration tools for musicologists and archivists
Business Impact
- Delayed digitization of priceless cultural archives
- Risk of losing traditional knowledge with aging gurus
- Inability to scale research across institutions
- Limited access for global researchers and students
How We Solved It
We built a comprehensive AI-driven platform that automates data collection from 10+ sources, cross-references information using multiple AI providers, and routes results through a human verification pipeline - creating the most accurate and scalable Indian Classical Music knowledge base ever assembled.
Multi-AI Data Extraction
Integrated OpenAI, Gemini, and Perplexity to extract and cross-validate musical data from diverse sources including web scraping, PDFs, and digital archives.
Human-AI Verification Loop
Built a collaborative workflow where AI-extracted data is routed to domain experts for validation, ensuring 98%+ accuracy on all published entries.
Structured Knowledge Graph
Designed a comprehensive data schema linking artists, raags, taals, compositions, and performances into a searchable, interconnected knowledge graph.
What It Does
Core capabilities that make this platform powerful and unique.
Multi-AI Integration
Leverages OpenAI, Gemini, and Perplexity simultaneously for comprehensive data extraction and cross-validation.
Intelligent Search
Advanced NLP-powered search across artists, raags, taals, and compositions with contextual suggestions.
Verification Pipeline
Human-in-the-loop verification system ensuring 98%+ accuracy on every published data entry.
Knowledge Repository
Structured storage of 10,000+ verified entries across multiple musical knowledge dimensions.
Web Scraping Engine
Automated scraping from 10+ data sources with intelligent deduplication and conflict resolution.
Collaborative Editing
Multi-user workflows with role-based access for researchers, verifiers, and administrators.
The Process
A step-by-step look at how the platform operates from input to output.
Data Sourcing
RagaBot scans 10+ sources - websites, digital archives, PDFs, and books - extracting raw musical knowledge.
AI Processing
Three AI providers cross-validate, structure, and reconcile extracted data into standardized entries.
Expert Verification
Domain experts review AI outputs, correct inaccuracies, and approve entries for publication.
Knowledge Graph
Verified data is linked into a rich, searchable knowledge graph connecting artists, raags, and more.
Tech Stack
The full technology stack powering this project, grouped by layer.
Frontend
Backend
Database
Cloud & DevOps
Integrations
Connected Platforms
External services and APIs powering this solution.
OpenAI
GPT-4 for data extraction and analysis
Google Gemini
Cross-validation and enrichment
Perplexity
Web-based research and fact-checking
AWS S3
Scalable document and media storage
At a Glance
Impact & Results
Measurable outcomes that demonstrate the real-world impact of this project.
Verified accuracy across all published entries through human-AI collaboration.
Reduced research time from weeks to hours per artist entry.
Automated scraping and extraction from diverse digital sources.
Verified musical metadata entries in the knowledge repository.
The People Behind It
Darshan Vasani
Project Lead & Full-Stack Developer
AI Research Team
AI Integration & Pipeline Design
Saptak Domain Experts
Musical Knowledge Verification
UI/UX Design Team
Interface & Experience Design
“RagaBot has transformed how we digitize and preserve our musical heritage. The AI-powered extraction combined with expert verification gives us confidence in every entry. This platform is truly one of a kind.”
See It in Action
RagaBot Dashboard - Main Interface