Sai Vijay Ragav

Developed a multimodal RAG pipeline using LangChain and Unstructured to ingest, parse, and index over 10,000 patent documents, adding metadata to improve downstream retrieval and analysis. Designed a hybrid search system backed by Milvus, supporting semantic retrieval across 1M+ embeddings and reducing query response times to under a second for internal knowledge discovery tools.

AI/ML Intern at Sprelia (2025-10 – 2025-12)

• Developed a multimodal RAG pipeline using LangChain and Unstructured to ingest, parse, and index over 10,000 patent documents, adding metadata to improve downstream retrieval and analysis.
• Designed a hybrid search system backed by Milvus, supporting semantic retrieval across 1M+ embeddings and reducing query response times to under a second for internal knowledge discovery tools.

Software Development Intern at National Informatics Centre (2025-05 – 2025-08)

• Built a Spring Boot ticketing platform integrated with FastAPI and Qwen2.5-VL to automate extraction of 8+ metadata fields, reducing ticket submission effort by 60–70%.
• Developed a RAG-based support assistant with PostgreSQL integration across 10+ ticket categories and workflows spanning 20+ states, accelerating issue resolution for support teams.

B.E. in Computer Science (Artificial Intelligence & Machine Learning) – Rajalakshmi Institute of Technology (2023 – 2027)

Diploma in Data Science – Indian Institute of Technology (2024 – 2026)