Dec 2020 – Jun 2022 · 1 yr 6 mos · Gurugram, Haryana, India

Jul 2020 – Nov 2020 · 4 mos · New Delhi, Delhi, India

Apr 2020 – Jul 2020 · 3 mos · New Delhi, Delhi, India

Jun 2019 – Jul 2019 · 1 mo · Greater Delhi Area

• Developed a backtesting framework to test trading strategies on historical data.

May 2019 – May 2019 · 0 mo · Gurugram, Haryana, India

• Member of the Order Entry Core Team.

May 2018 – Jul 2018 · 2 mos · Saarland

• The project's goal was to develop a tool which performs mixed-precision optimization fixed-point arithmetic. In particular, the main part of the project was to use machine learning to learn a suitable and efficient cost function, which is used during the optimization. This function needs to roughly approximate the behaviour of an FPGA compiler, for our target platform, as such a compiler is too expensive to be run during the optimization search itself. As the behaviour of such a compiler is very complex, machine learning appears to be a very good fit to generate an approximation. The project consisted of generating suitable data for learning, including a selection of features, and the evaluation of different machine learning algorithms. As a result, a cost function was found with a relative error below 5% (approximately). This work will be incorporated into the existing framework and likely published in the near future.