Hemang Shah is a Software Development Engineer II at Amazon, specializing in cutting-edge AI/ML-based solutions for IP protection and search optimization in large-scale cloud environments. With expertise in semantic search, computer vision, and NLP, Hemang has architected innovative systems that safeguard intellectual property across Amazon’s 20 billion+ product catalog. His accomplishments include designing embedding-based similarity search engines using models like CLIP and MoCoV2, achieving 85% accuracy in detecting counterfeit products, and removing over 14,000 infringing listings daily.
Hemang’s work also extends to optimizing multimodal vision-language models like InstructBLIP and BLIP2, enabling 80x throughput improvements in IP detection pipelines and scaling capacity to process 1.6 million products per day. By leveraging AWS technologies such as SageMaker, OpenSearch, and Step Functions, he has reduced manual auditing efforts by 80%, saving Amazon $6.75M annually. His passion for scalable and cost-effective architectures drives his success in integrating advanced ML capabilities into enterprise workflows.
Hemang holds a Master of Information Systems Management from Carnegie Mellon University and a Bachelor’s in Information Technology from SSN College of Engineering. A frequent contributor to research and innovation, he has presented at IEEE conferences and developed AI-driven systems that enhance operational efficiency across diverse domains. Hemang is committed to empowering organizations to leverage ML for IP protection and delivering actionable insights through data-driven strategies.