Jan 2022 – Nov 2023 · 1 yr 10 mos

• https://quantiacs.com/ competetion participant. Tried out various portfolio optimization techniques. Backtested and paper traded. The computation is done in python using libraries like riskfoliolib etc
• Made an Options pricing mini library using first principles in C++. Dependencies were libraries like Eigen, ML pack. The data used to reconcile was publicly available Indian options 1-minute bars for weekly and monthly expiries. Researching Systemic Option strategies to deploy.
• Various online courses on ODE, PDE, Stochastic Calculus, Numerical Analysis, and Convex optimization techniques. The code is present on GitHub.
• With a group of independent collaborators curating global macro strategies from an Indian market perspective. To be shared in some detail on substack soon.

Sep 2021 – Mar 2023 · 1 yr 6 mos

• Heading the Data Science team.
• Creating ML factor models for various frequencies for large institutions - family offices, hedge funds, CPOs.
• Creating high-frequency predictor PoCs for large institutions (market makers) via quant/ML literature research. Using models like optimized LightGBM, and specialized DNNs (CNN, RNN, Encoder-Decoder) for Limit Orderbook data and trade data. These pertained to Equity and Futures market making and involved large North American clients.
• backtested/simulated high Sharpe low drawdown derivative strategies for various cryptocurrency clients. Also, included TCA and market impact analysis. These were successfully released for multiple clients. The codebase was in C++ and python. Consistently profitable for most clients in paper and live trading.
• Supervising the creation of a real-time Feature Engineering pipeline. Involves curating factors/features from Quant Finance literature. Also, consists in dealing with the acquisition of data from various global vendors like Sharadar, Factset, Algoseek etc
• Working with the engineering team for the creation/maintenance of a common scalable and robust Financial machine learning pipeline. This includes preprocessing, explainability and modeling. Used technologies like PySpark, Dask, RAPIDs etc on Big Data platforms.
• Audited Corrective AI Forex strategies of myriad frequencies for key client fund.
• Involved in technical recruiting for Quant and Machine learning profiles.

Oct 2020 – Mar 2023 · 2 yrs 5 mos

• Working on Financial data science and ML infra.
• Managed the Quant Factor Engineering efforts.

Sep 2020 – Feb 2021 · 5 mos · Mumbai, Maharashtra, India

• Worked on Indian Commodity futures quant strategies.
• Used Quant Fin literature to create various hypotheses for cross-sectional momentum in commodities.
• The evaluation metrics were below stipulation so the strategy didn't go live.

Nov 2019 – Sep 2020 · 10 mos · Mumbai, Maharashtra, India

• Machine learning and financial research on various quant finance strategy paradigms. Back tested and paper traded.
• Mentored 7+ EPAT participants for Quant Machine learning projects.

Mar 2018 – Sep 2019 · 1 yr 6 mos · Bangaon Area, India

• The Natural Language Processing Lab.
• ● Worked with Prof. Deepa Gupta on the KCC ( Kisan Call Centre) government data, in collaboration with Tamil Nadu Agriculture University with an aim to create a semantic conscious FAQ system to automate response in Kisan Call Centres. The dataset had question-answer data pertaining to all districts from all farmer call centres across the nation.
• ● Used various existing cutting edge and novel deep learning models (RNN, CNN, Attention, Variational, Autoencoder) to create sentence embeddings using existing ​ word embeddings for semantic clustering (​ Paper in pre-print)
• ● Created various Reinforcement Learning methods in the domain of ​ Active learning and used the reduced training set for various downstream tasks. Further, conceptualised literature that we aim to publish.
• ● Received two publication acceptances from international conferences

Jun 2015 – May 2017 · 1 yr 11 mos · Bengaluru Area, India

• ● Was a member of the core KBB (Keyword Black Box) team of the online ads company Media.net.
• ● Developed, maintained the scalability, robustness and availability of the Keyword Black box API which was the core API and returned the ad keywords based on either the context of the page or the profile of the page viewer.
• ● Developed and maintained in-house monitoring/alerting tool - Slant.
• ● Deployed various tools for monitoring, scalability, metric collection, automation etc.
• ● Helped optimize DC server usage which led to​ cost-cutting​ of ~ ​ $50,000 ​ over a year.
• ● Conducted technical recruitment and evaluation process of freshers joining technical teams.