Saumya Gandhi

Saumya Gandhi is an emerging AI/ML engineer and researcher with a strong academic foundation from Carnegie Mellon University and Visvesvaraya National Institute of Technology. She specializes in large language models, natural language processing, and the development of high-quality synthetic data for AI applications. Her recent role at Anthropic as a Member of Technical Staff in the Applied AI Fine-tuning team, following her leadership in model quality at OpenPipe, highlights her dedication to advancing powerful and safe AI. github+2

Her research contributions include developing 'DataTune,' a novel method for generating retrieval-augmented synthetic data, and multiple publications on suicide ideation detection using transformer-based models and deep adversarial learning. This body of work demonstrates her expertise in practical applications of NLP and her commitment to creating robust and high-performing AI systems, particularly with a focus on societal impact in mental healthcare. github+2

Saumya has diverse experience across major tech companies and research institutions, including a Summer Analyst role at Goldman Sachs, a Research Assistant position at Oracle, and various internships focusing on AI and software development. Her early career also includes co-founding a debate club and mentoring students, showcasing leadership and a proactive approach to learning and community involvement. github+1