Sep 2023 – Feb 2026 · 2 yrs 5 mos · United States

• Applied Science role - Leading the Eval framework for agent outputs
• Leading the privacy reviews, security reviews and responsible AI reviews within my team
• Leading the development of AI agents in D365 Sales using Microsoft Copilot Studio
• Backend Engineering team responsible for building a core microservice for D365 Marketing.
• Reduced COGS by 10% amounting to over 4.5 million dollars.
• Winner of Dynamics 365 Marketing Hackathon 2023.
• Winner of Dynamics 365 Marketing - AI Sales Hackathon 2024.

Jul 2021 – Sep 2023 · 2 yrs 2 mos · United States

Jan 2021 – May 2021 · 4 mos

• Research intern with the SEAM lab at UMD.

Sep 2020 – Dec 2020 · 3 mos

• CMSC 424: Database Design

Aug 2019 – Sep 2020 · 1 yr 1 mo

• As a part of the SEAM lab, I am working on building a "Platform for structured discussions and peer review" under the guidance of Dr. Jim Purtilo. The project involves application of methods and techniques from Decision Making, Psychology and Human-Computer Interaction. The scale of the project also poses many engineering challenges.
• Team: https://seam.cs.umd.edu/team

May 2020 – Aug 2020 · 3 mos · Los Gatos, California, United States

• Working with the Customer Trust Team. I am helping the team identify signals for Account Takeover and implementing models to detect fraudulent events. Created a statistical model to automatically identify ever-changing rules to identify fraudsters.
• On the side, I am working on a library to improve developer productivity to help them code faster in their Big Data Infrastructure. The library will soon be integrated in Netflix's core data science platform and will benefit more than 500 developers.

Feb 2018 – Jul 2019 · 1 yr 5 mos · Bangalore

• Led a team of engineers to solve the problem of "Automated Discovery of Automatable processes" under the guidance of Dr. George Nychis and Shankar Natarajan. We applied this technology to measure human productivity while performing software related tasks and for finding automation opportunities.
• Built scalable and sophisticated tools, pipelines and methods to measure employee productivity and automation opportunities with high accuracy.
• Worked on building an analysis pipeline to process hundreds of millions of data points to discover repetitive sequences. Implemented and evaluated various clustering, data scrubbing and frequent pattern mining techniques on the collected datasets.
• Designed the system to be secure and GDPR compliant. Performed various tests using automated tools and did vulnerability analysis to ensure production-grade security. Used cryptography in a creative way to enable deleting of data (amounting to terabytes) within seconds of a request from the client.
• Built our own in-house deployment pipelines. Worked majorly on creating the entire end to end infrastructure to make this possible. The tool helped in setting up a server from scratch, building an environment for tools to run and managed automatic updates to more than 20 Microservices.
• Assisted various other teams in the organization by helping with brainstorming, code reviews, deciding the product roadmap and writing technical specifications.
• Helped the organization with interviews, setting recruitment tests, and by going on on-campus recruitment drives.

Jul 2016 – Feb 2018 · 1 yr 7 mos · Bangalore

May 2015 – Jul 2015 · 2 mos · Gurgaon, India

• Summer research intern under the guidance of Dr. Ehtesham Hasan and Dr. Swagat Kumar at the Multimedia and Robotics Lab, TCS Innovation Centre, Delhi.
• Worked on building an agent using LIDA (Learning Intelligent Distributed Agent), a
• biologically inspired cognitive architecture to solve the problem of SLAM (Simultaneous Localisation and Mapping). It was a unique research problem where an autonomous agent was modeled on the human reaction system.
• I was the only undergraduate student to receive this internship in 2015. Was awarded 4/4 for the performance.

May 2014 – Jul 2014 · 2 mos · Surat

• Worked on a project titled "Real-Time Facial Expression Recognition using Eigen Face Model" under the guidance of Dr. M.A.Zaveri.
• The project was successful as it managed to achieve an accuracy of 96%. A research paper on the same topic was presented at the Section of Information Technology and Computer Science, Indian Science Congress, the University of Mumbai in January 2015.