Jun 2023 – Nov 2024 · 1 yr 5 mos · New York, United States · Hybrid

• Working on generating alpha from alternative data and trading patterns.

Apr 2023 – Jun 2023 · 2 mos · San Diego, California, United States · On-site

• Teaching Assistant : Machine Learning (DSE220: Dr. Sanjoy Dasgupta)

Sep 2022 – Jan 2023 · 4 mos · San Diego Metropolitan Area

• Teaching assistant for course Design of Algorithms (CSE-202: Dr. Rammohan Paturi)

Jan 2023 – Jun 2023 · 5 mos · California, United States · On-site

• Working on recognizing the different T cells causing Asthma using ML techniques. (Advisor: Dr. Vijayanandan Pandurangan)

Jun 2022 – Sep 2022 · 3 mos · San Francisco Bay Area

• Build low latency ML pipeline for federated learning for AWS clients,
• Build Jupyter / IPython extensions on top of it for easier development.

Jan 2022 – Aug 2022 · 7 mos · San Diego, California, United States · On-site

• Efficiently identify sequencing errors using probability models which exploits the neighboring reads to process the false
• variant calls, to improve haplotype genome assembly. (Advisor: Dr. Vikas Bansal)

Jul 2018 – Jul 2021 · 3 yrs

• Redesign overlays used in strategy testing from file based system to key based system:
• Overlays are used for testing purposes, initially it copied files from the production system during its creation, causing
• stale data for future usages.
• Automated the update of cached data for strategy classes lazily per key, improving efficiency and robustness of the
• system.
• Created option to keep different types of file based systems (H5, Pkl, Redis backends) in the overlay as compared to
• the prod for better readability and efficient storage.
• Performed optimizations for efficient storage of data (ranging to ten years of historical data) and quick queries on it:
• Sharding on custom key axes, enumeration of sparse data, parallel chunked updates.
• Optimize aligning data with non-prefix keys using almost-dense vs sparse optimizations during cache retrieval.
• Automate sharing private data associated with strategy’s class instance, without exposing the private strategy code, making
• the data shareable much more easier within teams, for various testing and analysis purposes.

May 2017 – Jul 2017 · 2 mos · Magdeburg, Saxony-Anhalt, Germany

• Implement and test viability of different Model Order Reduction techniques used for dimensionality reduction on GPU.
• ◦ Implement the algorithms in CUDA, and test on several Linear Dynamical Systems, e.g., HEAT BEAM models to test its correctness and stability

Dec 2015 – Jan 2016 · 1 mo · New Delhi Area, India

• Optimized and enhanced internal search engine on top of Elasticsearch for:
• queries with typos and incomplete words and making it case insensitive
• return relevant results based on user ratings, number of views of the page and user feedback