Intern
Bharat Electronics Ltd.
- Contributed to a multimodal GenAI system for extracting and understanding complex PDF documents (tables, images, embedded spreadsheets) as part of an ML-driven data pipeline.
- Worked with Docling for structured document parsing, and Mineru for high-fidelity image extraction, enabling reliable preprocessing for downstream ML tasks.
- Assisted in developing a RAG-based architecture for context-aware question generation, producing Bloom’s Taxonomy–aligned question banks and answer keys.