Atharva Shrikant

About Atharva Shrikant

Seeking a challenging SDE roles where I get to showcase my skills in solving ground breaking problems with strong, agile and dynamic team. Deeply intrigued by various facets of computer science, I am a Software Engineer with overall 3+ years of experience working with highly distributed scalable systemsProficient in problem-solving and system design. Proven experience in Java, Spring Boot, Microservices, Angular, CI/CID.Involved in all parts of SDLC lifecycle. Always keen and updated on latest disruptive technologies.Keep up with my DS & A skills and High-Level and Low-Level Design.Diligent. Dedicated. Always

Experience

1. Software Development Engineer Ii

   Amazon

   Jan 2022 — Present · Bangalore Urban, IN

   Designed and led the implementation of the Reaction Archive (DLQ) for EventBus, a multi-tenant,high-throughput queueing system processing 10M events/sec across Sable/DynamoDB updates. This feature enabled safe handling of persistently failing messages, reducing reaction misses and improving system reliability. The improvement significantly reduced load time and latency on critical Amazon.com pages such as Search, Product Detail, and Checkout, enhancing customer experience at scale. Achieved $1M/year in compute and storage savings, and cut overall P99.99 propagation latency of precompute platform by 66%(from 47 mins to 16 mins). Reduced compute costs for customers by as high as 91%. Architected a region-agnostic, multi-cluster EventBus framework as part of Amazon’s org-wide Region-Flex initiative. Enabled EventBus to read live traffic from both DUB(Ireland) and ZAZ(Spain) regions and dynamically assignwork across workers. Unblocked traffic migration for 600 precompute applications and Sable scopes, improving power efficiencyand scalability. Developed operational readiness tooling and automation to support regional migration for EventBus and services like Precompute, BackfillService, and Sable, saving 100+ SDE hours per day and eliminating manual errors. Enabled successful migration of 10% of DUB traffic ( 400K events/sec) to ZAZ, potentially saving $6.2M in power consumption annually. Built internal self-service tooling for deferred queue management, allowing faster recovery for customers which reduced their bills by 63% MoM. Diagnosed root cause and resolved connection pooling issues for Middle Mile, a Tier-1 service. It reduced P99 propagation latency by 90%, achieving sub-second response times. This prevented disruption in assigning delivery trucks for shipment, maintained prime delivery promises and customer trust. This also avoided a potential order rate drop by 26% in EU.

Education

• Institute Of Engineering and Management

  Bachelor of Technology - BTech, Computer Science and Engineering

  2014 — 2018