Kashyap Dixit — Software Engineer

I have moved to ECS Scheduler team in AWS. We are responsible for maintaining long running ECS services, ensuring health and availability. I have many years of experience working in designing, planning and developing large scale distributed services in a variety of domains. In contrast, I have background in theoretical computer science (PhD) and mathematics. I am also an active reviewer of submissions to Mathematical Reviews (MathRev), a journal by American Mathematical Society. Earlier, I worked for Audible, in their personalization group. The worked involved development and maintenance/monitoring of high throughput low latency micro services, gathering customer feedback asynchronously through event streams (kinesis, sns/sqs) and using them for filtering and improving personalized recommendations for customers, enabling ML models based data publishing through automated pipelines. I have worked in search and personalization teams at jet/Walmart labs and Compass, where I had experience working with Elasticsearch, Kafka, docker and kubernetes, Sql and NoSql databases, Redis etc. for building high throughput and low latency ingestion pipelines and search services. I have completed my Ph.D from Penn State University in Computer Science. I have mainly worked on designing fast randomized algorithms for large inputs. This involves designing sublinear time algorithms for property testing and polynomial time efficient algorithms for counting patterns (subgraphs) in large graphs. Sublinear time algorithm for this problem will save a lot of pre-processing time for many applications in databases and machine learning. We have come up with such algorithms for many problems. My work has been published in reputed journals and peer reviewed conference proceedings (SODA, ICALP, TCC, CIKM etc.) My latest work is focused on counting complete subgraphs (cliques or closed-communities in social network parlance) in large input graphs. This work has resolved a problem that was open in computer science and maths community for decades. At IBM research, my work was focused on ranking and matching algorithms for workforce management related projects, which involved designing and implementing optimization algorithms for underlying matching and scheduling problems that came up in these projects. Apart from theoretical and mathematical background, I worked with JAVA and Python based machine learning and optimization APIs. I used optimization libraries like CPLEX and CVSOPT.

Stackforce AI infers this person is a Backend-heavy Fullstack Engineer with expertise in SaaS and E-commerce.

Location: Harrison, New Jersey, United States

Experience: 17 yrs 9 mos