Jun 2021 – Feb 2022 · 8 mos · Bangalore Urban, Karnataka, India

• Leading a backend team for the development of revamped Storefront platform which integrates with Fabric components.
• Led the architecture design and implementation of a modular Golang / GraphQL-based backend-for-frontend (BFF) service powering Fabric Storefront platform. This involved careful consideration of trade-offs while ensuring system robustness and extensibility.
• Productionized the backend service within 4 months with a lean team of 3 engineers by clarifying requirements, streamlining processes, managing stakeholder conversations and setting technical standards.
• Drastically improved provisioning of backend services at Fabric by creating a reusable infrastructure and CI / CD template using Terraform, AWS CLI and Gitlab CI. This became the reccommended way to create services at Fabric.
• Developed the first fully functional end-to-end checkout flow for e-commerce websites built using Fabric Storefront Platform (integrating with all Fabric components - Fabric OMS, Fabric Offers, Fabric PIM etc) by integrating modular Fabric APIs.
• Created the BFF technical roadmap by interfacing and achieving consensus with engineering and product leadership across Fabric.
• Actively participated in design reviews and provided consultation to other backend teams at Fabric.
• Provided technical guidance to the team by actively participating in code reviews and encouraged shared learning by holding Tech Talks on various useful topics.

Apr 2020 – May 2021 · 1 yr 1 mo

• Working in the Observability team. Specifically focusing on implementing distributed tracing across Grab's backend and improving the feasibility of high-cardinality monitoring.
• Created a solution to automatically alert developers when services generate metrics with High Cardinality which may result in billing increase for third-party metrics provider. This resulted in savings in-excess of US$xx,000 per month.
• Laid down a vision for observability and distributed tracing improvements at Grab - aligning long-term team vision with short-term project goals.
• Optimized existing infrastructure for third-party distributed tracing provider for better performance. Improved recall times by optimizing load-balancing for trace endpoints as well as modularized existing infrastructure using Terraform.

Apr 2019 – Mar 2020 · 11 mos

• Working in the Common Frameworks team. Working on solutions to help make monitoring and debugging easier at Grab as well as frameworks to accelerate feature development and build resilient backend services.
• Led a team of two to create a solution to automate monitoring for backend services, using monitoring-as-code. Integrated with in-house deployment automation to enable automated rollbacks of bad canary deployments. This enabled automated monitoring of 100+ previously unmonitored or poorly monitored services.
• Created a solution to minimize cross-AZ inter-service communication by overhauling client side load balancing logic at Grab, resulting in reduced network costs and increased savings.
• Enabled secure logging by creating a library to easily hide sensitive information from logs. Used to hide 600+ pieces within one month of release and has become recommended way to do secure log obfuscation at Grab.
• Implemented multiple resiliency improvements to shield backend services from bad network conditions. Notable ones include improvements to API retrying and improved health checks for connection recycling.
• Interviewed multiple candidates for recruiting engineers in the team. Also part of multiple hiring drives.
• Advised engineers on new projects outside team by providing suggestions on new RFCs.

May 2018 – Mar 2019 · 10 mos

• Working in the Common Frameworks team. Developing services, base frameworks and internal libraries powering most of Grab's services.
• Created a library for structured logging with features such as deterministic structure, dynamic log levels and customizable sinks / formats, vastly improving the debugging for backend services. Read more at https://engineering.grab.com/structured-logging.
• Enhanced developer productivity by envisioning easy-to-use features in Grab-Kit (in-house service framework) and Commons (diverse set of frameworks). Notable ones include speeding up code generation processes for large services by 20x and redesigning generated endpoint handlers for increased maintainability.
• Designed a vision around debugging and how to improve it, used to drive multiple goals across teams.
• Consistently part of on-call rotation. Mitigated and helped the team mitigate multiple high-impact production incidents.

Dec 2017 – May 2018 · 5 mos · Singapore

• Developer at EC department.
• Designed and developed a distributed architecture for generic spreadsheet-based bulk modifications and schema-free ETL from multiple data sources as a microservice, using Spark, Hadoop, Kafka and Cassandra. Used to develop all the major bulk features in the module. Awarded Star Performer by the director of the company for this effort.
• Designed and implemented a high speed bulk search re-indexer for e-commerce ERP using Spark and Elasticsearch bulk APIs.

Jan 2017 – Nov 2017 · 10 mos · Singapore

• Developer at EC Department.
• Developed UI/UX for e-commerce ERP using Google Closure and server-side rendering.
• Developed RESTful backends and APIs using Spring and JavaEE for microservice interactions.
• Developed a dynamic search UI/UX/Backend with all standard features (facets, autocomplete etc) using Elasticsearch.

Dec 2014 – Dec 2015 · 1 yr · Mumbai Area, India

• I worked on two projects with different professors at CSE, IIT Bombay,
• Interactive learning with Bilingual OCR. Designed and implemented a bidirectional LSTM with standard BTT for recognizing preprocessed bilingual characters.
• Improving Visual Attention using Deep Learning. Studied various methods for implementing attention using saliency and relative contrast. Implemented soft attention for ImageNet-1000 on top of a CNN.

Dec 2013 – Jan 2014 · 1 mo · Pune Area, India

• Designed and Implemented an optimization model in MATLAB for finding economically and thermodynamically most beneficial route for generating a required chemical product, given a certain set of reactants. This model also allowed specifying external conditions, such as separability, temperature etc.