About

Harsh Desai is an experienced engineering manager with over 11 years of experience in building, maintaining, and scaling distributed systems. He is currently working at Uber as an Engineering Manager, where he is responsible for designing architecture for a server-driven session management system that manages 30 million sessions between different LoBs for Uber riders, supporting 50000 rps. He leads a team of six engineers of Backend, Frontend, and data engineers to build, deploy, and migrate. He also orchestrated migration to a new session definition for thousands of datasets seamlessly, providing accurate business metrics improving session coverage to 99% percent of events, fares, and trips, and increasing reusability of signals and different ML models across different verticals. Harsh improved the real-time state management of the system, which powers 100 plus ML, Ops, and BI use cases. This needed refactoring 100K lines of code for better event management and complete change of architecture. This improved computing power by 50% by changing sharding and pipelines. He is currently working on a futuristic data platform model that can support different impressions, components and provide real-time user flow analysis. He is restructuring data consumption by different derived and raw tables and collaborating with multiple teams from engineering and Data Science on common logging requirements. This provides signals for Uber's super-app and personalisation use cases. Harsh has helped in growing a team of 10 engineers to 50 engineers by actively participating in hiring, requirement gathering, creating job descriptions, and actively engaging with HR partners. He has set up a vision and north-star for the team and collaborated with various teams and verticals for gathering requirements and roadmap planning. He is leading a team of 10 highly motivated, high performing engineers. He improved team satisfaction by 10% and developer velocity by 15% by taking various measures such as mentoring, peer programming, clarity towards roadmap, deep dive opportunity with the leadership team, one-to-one meeting every 2 weeks, and providing timely crisp and useful feedback. Harsh has previously worked at Qubole as an Engineering Manager, where he led the Platform team responsible for building scalable, fault-tolerant applications to support exponentially increasing customer growth. He conceptualized the roadmap for the team and executed it to perfection by using agile methodology. He planned and managed complex platform migration like platform movement to Kubernetes, control plane migrations between AWS to Azure with minimum to null impact to customers. He was involved in product design on both usability and technical