Summary
Vibhu Agrawal is an applied scientist with 8 years of experience building production-scale AI systems across multimodal LLMs, document understanding, computer vision, and medical imaging. At Amazon he works on agents for workflow automation and previously led deep learning at Nanonets, where he architected layout-aware language models and a constrained decoding framework to guarantee structured, non-hallucinated JSON outputs with spatially grounded key-value bounding boxes. He combines research pedigree—including a published medical imaging paper and a filed CT-scan patent—with hands-on engineering of scalable finetuning and adapter-based serving pipelines. Comfortable across encoder-only, encoder-decoder, and decoder-only model families, he has practical expertise in making LLMs layout-aware and efficient via techniques like LoRA. Based in Bengaluru, he is passionate about translating cutting-edge multimodal research into reliable, deployable systems that solve real-world document and imaging problems.
8 years of coding experience
5 years of employment as a software developer
BITS Pilani, Birla Institute of Technology and Science
Higher Secondary Examination (Indian School Certificate), 96%, Higher Secondary Examination (Indian School Certificate), 96% at Summer Valley School