San Jose, California, United States
Developing and Managing complex software applications. I have managed teams and developed complex applications within Search backend team for eBay. The team I managed is responsible for indexing/query serving infrastructure for search results data to eBay users. I have managed and led various solutions to Search problems for eBay. I have collaborated with several teams(Product Management, Business, Operations) to make lot of initiatives possible. Even when I have been managing teams, I had hands-on experience and applied my engineering skills to solve various problems. At eBay I developed a snippets service after switching to a individual contributor role. I worked as a technical lead for Ad Serving group in Apple Inc. I am currently working as a tech lead in Amazon. I have been directly involved in building/architecting systems with the following technologies: - Low latency/ high throughput systems - Search engines/information retrieval - Ranking algorithms/Predictive models usingMachine learning - Various iterations of evolving technology for impression tracking infrastructure, real time search engines, large scale data processing systems with and without hadoop -- NO SQL systems -- Cloud based software products (Elastic Map Reduce) I also believe in continuous learning and have done lot of online courses to learn/improve my technical knowledge.
At Grainite, we are building a radically simple real-time application platform. By leveraging Grainite, any developer can easily build distributed applications by focusing just on their business logic. Stitching together a complex system with multiple components, to build streaming data applications, will be a thing of the past. Specifically I worked on the following: 1. Designed and implemented security architecture by implementing encryption at rest and encryption in transit. 2. Designed and implemented online backup/restore. 3. Implemented indexing for Distributed Replicated log used as message store and database changes. 4. Various performance improvements for reducing end to end latency.
Woked on Online Deep Learning Inference Engine for Query Understanding Intent Classifiers as part of Search group. Also worked in AWS Elastic Map Reduce team in Palo Alto enhancing Developer Experience tools as part of EMR Studio team.
I played a key role as a tech lead to develop a new ad server for Apple News application from scratch. I was tech lead for Non guaranteed delivery of display ads. My team developed models to predict click-through rate of ads .
After a few years in management, switched back to individual contributor position. In this position, I developed an automatic query-independent item summarization from description for ebay items. I also worked with development teams to make sure the transition from an old search engine to a modern one seamless.
In this position, I managed the software development for core search engine for ebay. I also had an indirect reporting structure that develops the feeder infrastructure for the search engine at our China development center. I was also responsible for development/deployment process. The work done by team generates multiple patents every year. My team worked with product managers/architects in several groups on an ongoing basis to implement/advise on Search implications and Search features. My team worked with Operations to create software solutions to reduce hardware, maintenance costs. Here are some of the recent projects that my team works/worked on. 1. Compression of the index to reduce the size of the index, 2. Image Search based on a pivot image 3. Distributed index creation using map/reduce 4. Supporting core search engine to be able to use models generated using machine learning 5. Supporting/understanding fitment queries for parts and accessories 6. Site speed initiative to improve the speed of SRP. 7. Created an impression tracking infrastructure that is available in Search. 8. Implementation of a virtual machine in search engine so we could execute dynamically created ranking functions. My team worked on implementing a significant portion of Cassini search engine which was a complete rewrite of a previous search engine all the while supporting various business/search science teams requests for features in the old search engine.