Sep 2023 – Jan 2025 · 1 yr 4 mos · Bengaluru, Karnataka, India · Remote

• Opsgenie Incident Management team
• Feature lead for e2e design, implementation, delivery & testing for JSM Incident Management Public APIs that contributes towards Opsgenie & JSM functional feature parity & overall consolidation.
• Added additional support for OAuth for the entire API suite enabling integration with Forge & Connect ecosystem
• Owned & delivered various CTB(change the business) features like increasing service limit support upto 20K(from 4K) unique services for customers like Sony, iFoods, etc. while ensuring overall system stability & performance.
• Feature lead for SLO audit & calibration of the entire suite of 20+ services owned by the Incident management team
• Worked on various RTB(run the business) tasks that are aimed towards improving team's operational posture in terms of efficiency wins like reconfiguring DLQ redrive configurations that have reduced manual requeue efforts by upto ~5X, implementing in-memory caches to reduce hourly downstream service calls by ~75%(reduction from ~1M calls to ~250K calls). Have also tackled various recurring issues that have contributed to reducing oncall load & overall alert count, like CPU/memory alerts, adopting Spinnaker for CI/CD etc.
• Co-lead a initiative to bring down production alert count for the team to increase alert fidelity & reduce overall on-call fatigue. ~85% overall alert reduction over a 6 month window from 4500 -> 600; 70% P1/P2/P3 alert reduction. Notable wins were reducing SQS related alerts by increasing queue throughput with the same host counts(using higher concurrent execution threads), reducing 4XX & 5XX exceptions, DynamoDB throttles, etc.
• Contributed to various company wide Cloudshift initiatives like Fedramp, BYOK, FIPS compliance, UGC log audit etc.
• Owned & contributed to various major version framework upgrades like Spring Boot migration(SBT 2.x -> 3.x), AWS SDK(1.x -> 2.x) upgrade & several security VULN patching initiatives

Mar 2023 – Aug 2023 · 5 mos

• Joined Alexa Automotive org as a part of ASV, Amazon Smart Vehicles.
• Currently working on the ingestion pipeline and serving layer APIs of an analytics platform for Alexa metrics used by both Amazon internal and external automotive partners and OEMs, to ensure GDPR compliance for various partners like BMW, Stellantis, etc.

Dec 2021 – Mar 2023 · 1 yr 3 mos

• Designed, & implemented from scratch, the core computation layer powering an interactive BI platform, with automated data ingestion & ETL capabilities, for DC Area Managers. It is built in Native AWS. It performs periodic data ingestion from a Data Lake, transformation in AWS Glue(Scala Spark) & persistence to DynamoDB. This is orchestrated with AWS Cloudwatch, AWS Eventbridge, AWS StepFunctions, AWS Lambda, & SQS. The layer also supports events from APIs in an EC2 service to perform incremental re-computation on user entered data. SNS is used for transactional reconciliation.
• The system generates various future projections on volume, rates, headcount & capacity delta related KPIs for forward looking Labor Planning exercises in DC sites & warehouses.
• Built a generic load-testing framework for all the systems involved in the tool & helping guide other SDE1s to deliver incremental feature additions to the tool. Have mentored multiple SDE interns to convert to FTEs & helped onboard multiple new hire SDEs to team tech stacks.
• Working on designing & implementing GraphQL APIs over OpenSearch indices to vend various KPI metrics in a real time fashion(as opposed to a batched ingestion)

Jul 2021 – Nov 2021 · 4 mos

• Amazon DCT (Distribution Centre Technology) Labor Management team
• Have done some integration PoCs with early stage AuthN & AuthZ platforms for RBAC.
• Have created a working optimiser model using MINLP solvers, using AWS Sagemaker & Docker, to orchestrate a PoC to allocate intra week headcount in manner that optimises capacity deficit, which provides a 6% improvement over the current GRG Nonlinear solve used by area managers in MS Excel

Jul 2019 – Jun 2021 · 1 yr 11 mos

• Amazon Access Points team
• Have worked in designing, scaling and implementing tech solutions to drive Access Point adoption and engagement across all marketplaces in Amazon Retail.
• Have worked on Perl based, legacy Amazon-proprietary MVC UI frameworks to instrument customer session metrics for A/B testing, from client-side JavaScript.
• Have worked in designing, and implementing changes in REST Web APIs, have learned some interesting working principles of customer traffic facing distributed systems, have owned the design & implementation of integration of team internal systems with a Tier 1 Retail PreCheckout service, scaling against upto 300K TPS of Pre Checkout traffic while maintaining a p99 of sub 5 ms incremental latency.
• Have gained some experience in operational excellence, software lifecycle management, real time incident response, service stress testing, & horizontally scaling services for high velocity peak events like Prime Day, BF/CM sales, etc. Carried out an optimisation that resulted in a 20X increase in availability of an Amazon-wide caching platform for hot records.
• Have carried out end to end stack migrations for various legacy systems powering critical customer facing UI components, including one where a precompute based O(n) space solution was migrated to a realtime compute based O(1) space solution, that brings in annual savings of ~100,000$ in infrastructure billing over data storage.

Sep 2018 – May 2019 · 8 mos · Kolkata, West Bengal, India

• Tasks involved assisting in coordinating campus placements & reaching out to potential recruiters, and invite them for campus recruitment.

May 2018 – Jul 2018 · 2 mos · Bengaluru, Karnataka, India · On-site

• I worked in Amazon Access Points team, and was involved in projects improving the direct delivery experience of end customers of the retail website, via changes made in the UX of the mobile app.
• I also made an internal API to add functionality to our team's service. The API involved writing and fetching data to and fro from a low latency non-relational database, while talking to an intermediate L0 caching layer.
• Technologies used: Perl Mason, Java Spring, Google Guava, JUnit, Mockito, TestNG
• Both the changes have been pushed to production, and subsequently the features have gone live.
• Received a return offer as a SDE, at the end of my internship.

Jan 2018 – May 2019 · 1 yr 4 mos · Greater Kolkata Area

• Involved in promoting awareness among students regarding the nitty gritties of tech interviews, and providing them exposure to real life interview scenarios via InterviewBit platform.