Manage ML-Infra training optimization team

• Optimized Nuro's models to improve training speed

• TL of training optimization team

• Technical owner of ML Lifecycle Management Platform(MLLP). Superset of MMS described below. MLLP is a core infrastructure that provides machine learning batch jobs, machine learning synchronous interaction and machine learning asynchronous interaction. Platform facilitates interaction with models in any manner with any machine learning framework.

• MLLP Platform manages thousands of jobs, hundreds of models and 7+ Adobe ML product services for enterprises

• Tech Lead for 3 - 4 engineers

• Technical owner of Model Management System (MMS). MMS is a multi-cloud, multi-language, multi-vendor internal service to manage the lifecycle of machine learning models at scale. MMS is a core infrastructure that helps engineers and data scientist build models from inception to production. This infra helped launch three major services and counting from beta to general availability for Adobe

• One of the first few engineers in a new organization within Adobe chartered to build machine learning services at scale. Implemented the initial architecture for our UI stack which was used to launch two AI services (Attribution AI, Customer AI). https://www.adobe.com/sensei/services.html

• Oversee development of other internal services of 1-2 engineers to help aid the feature requests coming from Adobe customers