Summary
Ashish Sardana is a deep learning solutions architect and engineer with nine years of experience building and deploying NLP and computer vision systems, currently a Member of Technical Staff at NVIDIA in the Bay Area. He combines hands-on research (from training 100M–1B transformer models and neural retrievers to developing custom speaker-embedding summarizers) with production expertise in scaling training across 1,000 GPUs and optimizing models for edge devices. His recent work spans hallucination detection for fine-tuned LLMs, multimodal document chunking to improve retrievers, and practical OSS integrations to help teams evaluate LLM reliability. Earlier roles include advancing ASR for Indic languages, medical imaging classifiers, and a prize-winning bionic prosthetic prototype, reflecting a pattern of applying ML to real-world, high-stakes problems. Comfortable moving between R&D and solution architecture, he also tinkers publicly with deep learning on GitHub and focuses on making research reproducible and production-ready.
9 years of coding experience
7 years of employment as a software developer
San José State University
Bachelor of Technology (B.Tech.), Electrical and Electronics Engineering, Bachelor of Technology (B.Tech.), Electrical and Electronics Engineering at Vellore Institute of Technology
High School, Computer Science, High School, Computer Science at Apeejay School
English, Hindi, French