1. Develop and fine-tune AI/ML models for the Shasan AI platform, including LLM integration using RAG architecture 2. Build and optimize NLP pipelines for processing government scheme data in multiple Indian regional languages 3. Work with local LLMs (Llama, Qwen) via Ollama and integrate with the Supabase vector database for citizen queries 4. Assist in building an intelligent chatbot and a scheme recommendation engine for citizen services 5. Conduct model benchmarking, prompt engineering, and performance optimization for production deployment 6. Collaborate with the core team on multi-agent AI workflows and API integrations