Nov 2023 – Present · 2 yrs 6 mos · Seattle, Washington, United States

• Leading the Digital Identity group to develop an accurate, safe, secure and compliant virtual representation of a customer's identity. My group owns multiple mission critical tier-0 access management / human relationship solutions that power SoFi products and enable new commercial offerings. Leveraging expertise in customer identity data, entitlement management, and authorization (FGA) to deliver a seamless member experience.
• We're just getting started. Ping me if you're interested in joining us and help customers get their money right!

Jan 2021 – Nov 2023 · 2 yrs 10 mos · Seattle, Washington, United States

• Leading multiple workstreams across the product and platform verticals to unlock financial freedom for everyone. My group is responsible for defining the future of Open Banking in North America, while enabling Plaid to seamlessly enable billions of connections on top of 11000+ financial institutions.

Jan 2020 – Jan 2022 · 2 yrs · Seattle, Washington, United States

Feb 2018 – Aug 2020 · 2 yrs 6 mos · Greater Seattle Area

• Strategic Initiatives, AWS
• Designed and led the implementation of the Data Disclosure framework on AWS to share highly sensitive healthcare data with authorized stakeholders in a secure and compliant manner in a time-bound fashion.
• Built Amazon Care's data lake/warehousing solution to store critical healthcare data of all the patients across the country.
• Launched multiple highly scalable microservices from the ground up with a focus on streamlined ingestion, encryption, deduplication, and data de-identification.
• Built a custom solution to solve the AWS Lambda cold-start problem leading to 49% improvements in latency.

Jan 2017 – Jan 2017 · 0 mo · San Francisco Bay Area

• 1. Implemented an alternate machine-learning approach for spelling correction at Amazon Search
• 2. Implemented a novel data-backed mechanism for assisting search at Amazon, simultaneously working on multiple codebases
• 3. Developed the end-to-end ML pipeline and a robust evaluation framework

Jan 2014 – Jan 2018 · 4 yrs

• 1. Owned the Legal Hold and the eDiscovery products, participated in architectural planning, implemented features to be rolled-out for biweekly sprint cycles
• 2. Fixed security vulnerabilities, redesigned UI, fixed optimization bugs and regressions over multiple products
• 3. Promoted to the role of Software Specialist and mentored incoming hires for quick-ramp up on the codebase

Jan 2013 – Jan 2013 · 0 mo

• I developed a speech recognition system using the new KALDI toolkit. I specifically focussed on implementing a Keyword Spotting System(KWS) in the minimum possible time. The complete code was written in C++. Apart from this, I worked on improving the recognition benchmark using the htk model on Malay language. I was also responsible for explaining the core concepts of the KALDI toolkit in weekly presentations, as well as exploring new ways of speeding up the system.

Jan 2012 – Jan 2013 · 1 yr

• Developed a real time speech recognition system using the htk toolkit. I worked on Hindi language and used HMM modelling to attain an accuracy of 67.8% on noisy data. I also implemented Artificial Neural Networks on top of that to achieve an accuracy of 74%.

Jan 2012 – Jan 2012 · 0 mo

• I worked on Computer Vision application to Pedestrian Detection and Tracking. I focussed my efforts on improving the hit rate for cases with occlusions and low resolution videos, working on publicly available PASCAL VOC2012 Challenge dataset and the large Caltech Pedestrian Dataset. Different techniques for tracking were explored and finally choosing Kalman filter for real time tracking.