Aug 2024 – May 2025 · 9 mos · Menlo Park, California, United States

• Llama Research Platform | AI foundations
• Building Meta’s RLHF platform
• Built a distributed training framework using torch distributed, reducing RLHF training time by 80%,
• significantly lowering GPU costs and enabling faster iterations.
• Enhanced the reliability of the offline inference package by integrating tests into the build pipeline, ensuring stable
• releases and reducing customer inquiries by 65%

Jan 2023 – Aug 2024 · 1 yr 7 mos · Menlo Park, California, United States

• Digital Commerce | Meta Fintech
• Building Meta’s Subscriptions platform
• Notable products launched
• Meta Verified, Ads free, Creator Subscriptions on IG and FB
• Founding engineer of Meta’s subscription platform; launched Meta Verified, Ads-Free Subs and Instagram Creator
• Subs, scaling to 4M+ active subscriptions and $750M+ in ARR.
• Migrated sync notifications to pub/sub architecture to improve reliability from 90.5% to 99.9%.
• Led response and resolution for 25+ service outages, driving swift recovery and long-term reliability
• improvements
• Led migration of data storage for subscription tables, improving read latency by 3x and write latency by 2x and
• enabling easier data model evolution.
• Optimized the post-purchase fulfillment endpoint by profiling performance, identifying bottlenecks, and
• implementing targeted fixes, achieving a 60% reduction in latency.
• Recognized with the ’Fixer’ award by the org leader for resolving longstanding platform issues and leading
• multiple service outages to uphold reliability standards.

Sep 2022 – Dec 2022 · 3 mos · Minneapolis, Minnesota, United States

May 2022 – Aug 2022 · 3 mos · Cambridge, Massachusetts, United States

• Data Lineage Infra | Data Infra
• Built a service to validate and stream lineage events from a queueing system to a thrift service. Currently handling 4B edges a day.
• Productionized the service by adding monitoring capabilities, enabling CI CD and fine tuning using CPU and memory profilers.
• When compared to the previous solution the new service was 15x faster and had 3x less memory footprint.
• Fun experience learning production C++, working on distributed systems at Meta’s scale and collaborating with cross functional partners.

Mar 2021 – Dec 2022 · 1 yr 9 mos · Minneapolis, Minnesota, United States

• CSCI 3081W - Program Design and Development (Spring 2021, Fall 2021)
• CSCI 4061 - Operating Systems(Spring 2022)
• MSBA 6331 - Big Data Analytics (Fall 2022)
• MABA 6490 - Machine Learning System Design (Fall 2022)
• Conducted office hours, led lab sessions, designed assignments, and wrote starter code for the course project.

Jan 2021 – Jan 2021 · 0 mo

Jul 2018 – Dec 2020 · 2 yrs 5 mos

• Designed and built a complete end-to-end flow of data ingestion modules for a utility containing 2 million users.
• Re-engineered the email creation framework to make it modular which in turn reduced the time taken to create a new email from ideation to production by 75%.
• Developed multiple public RESTful APIs using the Spring framework for utilities to use.
• Led a team of developers and testers responsible for setting up our suite of products with numerous custom features for a new client.
• Conducted technical talks on our product’s architecture and the working of different modules as part of the onboarding process for more than 30 employees.
• Developed numerous python scripts to automate tedious manual work for both developers and testers.

Jan 2018 – Apr 2018 · 3 mos · Bangalore

• Developed a pipeline monitoring system responsible for extracting time series data from AWS cloudwatch, storing it in InfluxDB, and providing intuitive dashboards
• Dashboards created with Grafana were more granular with better visualization making them easier to understand

May 2017 – Jul 2017 · 2 mos · Bangalore

• Contributed to the overall quality improvement of the architecture compliance kit
• Learned about the ARM architecture and the architecture compliance kit, found error-prone test cases, and resolved them. Integrated a tool in the kit to avoid such mistakes in the future

May 2016 – Jul 2016 · 2 mos · Bangalore

• Built an XLSX engine in Delphi (Object Pascal) to parse Microsoft Excel Documents. This was a part of the multi-purpose reader developed by Gnostice, which was capable of reading multiple file formats like .pdf, .jpeg, .docx, .xlsx, .pptx, etc