Jun 2022 – Aug 2022 · 2 mos · Bala-Cynwyd, Pennsylvania, United States

Jun 2021 – Aug 2021 · 2 mos · Menlo Park, California, United States

• Designed a fully automated NNPI test framework using Facebook internal tools to reduce testing time from one week to two days.
• Created a Facebook workplace bot to make posts about the success of NNPI tests to facilitate collaboration and information sharing
• Accelerated Facebook's file-sharing tool by a factor of 5. Expanded it further to check ShaSum} before downloading for better security.

Jan 2021 – Jun 2021 · 5 mos

• Devised first-ever model to address multi data and label owners, achieving 97\% accuracy for MNIST. Published work at CIKM'2021.
• Investigated momentum-based optimizers for non-iid datasets to accelerate convergence and enhance accuracy by 5%.

Jan 2020 – Jun 2021 · 1 yr 5 mos

• Developed a new neural model called OccamNet that can find interpretable, compact, and sparse solutions for fitting data.
• The model fits a variety of algorithms, ranging from simple analytic functions to simple image classification.

Aug 2019 – Jan 2020 · 5 mos

• Developed a graph network to classify the topological insulator correctly.
• Classified the large data on Topological Insulator and identified some incorrectly classified ones, which can then be used as a training set for future research
• Improved the RNN model by increasing the layers and making it immune to change in the order in which atoms were fed in the model.

Jun 2020 – Aug 2020 · 2 mos · Boston, Massachusetts, United States

• Created cyber security workflow to detect system attacks by using unsupervised machine learning algorithm on system logs.
• Monitored ten million logs using specialized tools like Splunk and Kibana to identify signals associated with affected systems.

May 2019 – Jul 2019 · 2 mos · Mumbai Area, India

• Worked with Apache, MariaDB, and application backend to fix various bugs and make improvements in the app.
• Also implemented a backup feature in case application crashes

May 2019 – Jul 2019 · 2 mos · Mumbai Area, India

• Developed a model to engage students in small conversations using speech recognition, tensor-flow library, and regex.
• Validated the model with real-time data and scored using Speechace API.
• Also developed games like scrabble using Tkinter GUI to make the bot engaging.