Building Distributed Services for Acropolis Hypervisor (AHV) Control Plane.

• Built SWW 2.0 Billing system. SWW - Ship with Walmart, is a multi-tenant SaaS product for providing shipping services with 3rd party carriers.• Involved in end to end product development from requirements gathering, design (LLD) and coding.• Made the complete pipeline Highly Available through multi-region data replication, multi-region Crons and Kafka failover.• Built a solid Reconciliation System to tackle the challenges associated with a distributed system as well as ensure the… Show more • Built SWW 2.0 Billing system. SWW - Ship with Walmart, is a multi-tenant SaaS product for providing shipping services with 3rd party carriers.• Involved in end to end product development from requirements gathering, design (LLD) and coding.• Made the complete pipeline Highly Available through multi-region data replication, multi-region Crons and Kafka failover.• Built a solid Reconciliation System to tackle the challenges associated with a distributed system as well as ensure the accuracy of transactions processed through the Billing System. Show less

• Developed a GraphQL BFF for an application to onboard sellers from US to Canada. Integrated with UI through API Gateway.• Wrote a Spark Job for generating reports to help sellers price their products profitably in the Mexico market.• Worked on the Catalog Porting feature to help sellers port their catalogs between international marketplaces. A catalog could involve millions of items in some cases and the feature was scalable in that sense.

‣ Worked on providing Number Plate Recognition as a service.‣ Built a distributed pipeline for serving Deep Learning models to several video streams in parallel.‣ Solved the challenges of Object Tracking, Stateful Processing and implemented Distributed Transactions using Optimistic Locking.‣ Technologies: ◦ Used Kafka as Message Queue and for Stream Processing ◦ gRPC for communication between Kafka and Model servers ◦ Protocol Buffers for serialization ◦… Show more ‣ Worked on providing Number Plate Recognition as a service.‣ Built a distributed pipeline for serving Deep Learning models to several video streams in parallel.‣ Solved the challenges of Object Tracking, Stateful Processing and implemented Distributed Transactions using Optimistic Locking.‣ Technologies: ◦ Used Kafka as Message Queue and for Stream Processing ◦ gRPC for communication between Kafka and Model servers ◦ Protocol Buffers for serialization ◦ Redis as a Distributed Cache ◦ Docker for Containerization and ◦ MySQL for persisting the extracted data Show less

‣ Interned in the Advertisement Technology team. ‣ Created a mechanism to report which ad campaigns must be paused or resumed, depending on the item availability.‣ Parallelized database queries and batched requests to get optimal performance while staying under the QPS limit.‣ It was deployed to production. ‣ Technologies: Spring Boot, Hibernate, Thymeleaf, MySQL

‣ Worked with the New Initiatives team.‣ Developed a solution for detecting if a worker is wearing safety gear while working at the workshop, using CCTV cameras.‣ Used YOLO for object detection and Deep SORT for person tracking.‣ Technologies: Python, TensorFlow, Tkinter (for UI)Code: https://github.com/AnshulSood11/PPE-Detection-YOLO-Deep_SORT