Work 👨‍💻

NYU Agile Robotics and Perception Lab

I am a researcher at NYU ARPL, where I focus on developing multi-robot systems using innovative techniques. Additionally, I have contributed to creating reproducible builds for the lab's pipeline.

NYU Building Informatics and Visualization Lab

In my role as a machine learning engineer at NYU BiLAB, I did some data annotation for the development of a facade defects detection and image generation project. I also experimented with Generative Adversarial Networks (GANs) and image detection models. The objective was to generate synthetic facade images while also segmenting various types of defects. Our efforts, earned us 3rd Place Prize in the NYU Tandon Research Excellence Exhibit, 2024.

MindsDB

During Hacktoberfest 2023, I integrated nine APIs, including PyCaret, Pinecone, NPM, SAP ERP, Lightdash, Xata, Milvus, Apache Doris, and Shopify, into the MindsDB ecosystem. This experience broadened my technical proficiency and also provided me insights into the intricacies of API integration within a compressed time frame.

Central Road Research Institute of India

As a full-stack engineering at CRRI, I developed web applications focused on processing raw data sets. These applications were designed to offer an intuitive user interface for data visualization and report generation. The projects also entailed significant geospatial data processing components.

Purdue University - CS252

I worked as undergraduate teaching assistant for systems programming course at Purdue University. The course was centered on the development of low-level software like malloc and shell in the C programming language.

Cummins Inc.

During my brief term as software engineer intern at Cummins, I wrote a service in Python using AWS Lambda and DynamoDB. I also implemented a developer SDK in Java to be used within the company. Additionally, I won 3rd Place Prize in company wide intern hackathon with a group of coworkers.

Publications 📑

Evaluating GAN Architectures for Generating Images of Defective Facades

We evaluated the performance of DCGAN, BigGAN, and Lightweight GAN architectures in generating facade images featuring four types of defects: erosion, efflorescence, missing mortar, and spalling. Our primary goal was to generate additional data to augment existing datasets. Among the models tested, Lightweight GAN proved to be particularly effective for this application.

Education 🎒

New York University

Master of Science: Computer Science

Purdue University

Bachelor of Science: Computer Science, Minor in Business Economics, Certificate of Entrepreneurship