Jul 2020 – Mar 2024 · 3 yrs 8 mos · India

May 2019 – Jul 2019 · 2 mos

• Interned with the Limestone trading team.
• Built ML and statistical models to predict market signals, used to generate buy-sell triggers for an aggressive trading strategy.
• Employed a variety of feature selection and data sampling techniques to efficiently use large historical market data.
• Experimented with Neural models and ensembling to improve PnL scores.

Dec 2018 – Jan 2019 · 1 mo

• Interned with the team responsible for task allocation and task prediction systems.
• Developed a horizontally scalable Geohash based, storage architecture which impacts 1M orders/month. Used consistent hashing for load balancing.
• Implemented concurrent BFS over geo hashes for querying all runners and all queued tasks, used the Hungarian algorithm for task-runner matching.
• Leveraged knowledge in concurrent software design to improve performance.

May 2018 – Aug 2018 · 3 mos

• Added SOTA rational point finding algorithms for curves and schemes in Sage.
• Reduced complexity of existing enumeration algorithm by a factor of d^5 (dimension), using number theoretic optimizations and advanced data structures.
• On a typical benchmark test, there was a reduction from 24.82 sec to 0.75 sec.
• Was included in the sage developer map (amongst 271 major sage developers in the world), for contribution in rational point algorithms.

May 2017 – Jul 2017 · 2 mos

• Worked on a scalable web application with an extensive technology stack.
• Explored real-time communications using WebRTC APIs.
• Integrated search component in the app using Elasticsearch.