Abhay Singh Thakur
I recently earned my Master's in Computer Science at Rutgers (May 2026), specializing in AI and machine learning. I lectured the ML Principles course, did graduate research in 3D object detection, and ship my own products on the side.
Who I am
I recently completed my Master's in Computer Science at Rutgers University (May 2026), specializing in AI and Machine Learning. At Rutgers I lectured the undergraduate Machine Learning Principles course and worked on graduate research in 3D object detection, fusing LIDAR and camera sensor data. Before that, I earned my B.S. in Computer Science and Data Science at Purdue University, where I built backend services used by 10,000+ students and taught Data Structures and Algorithms as an undergraduate TA. I have also interned at Pacific Northwest National Lab, automating infrastructure for security research testbeds with Terraform and Ansible. Outside of coursework and research, I ship my own products - including a Flutter app live on the Apple App Store - and build full-stack and ML projects ranging from RAG-based medication safety tools to deep learning pipelines for sports video. Hackathons got me started (1st place at Purdue Hackers' Hackathon, top 10 at BoilerMake VIII), and that habit of building things end-to-end has stuck. I am currently looking for software engineering and machine learning roles where I can keep doing exactly that.
Work I've done
- Engineered a high-throughput pipeline for 3D object detection fusing LIDAR and camera data, improving localization accuracy 25% over baseline.
- Built backend modules that auto-generate 3D bounding-box metadata from raw sensor inputs, cutting manual annotation ~40%.
- Implemented geometric transformation services with 98% reprojection accuracy using configurable camera intrinsics/extrinsics.
- Lecture the undergraduate Machine Learning Principles course - supervised learning, model evaluation, neural networks.
- Run weekly office hours and help sessions, improving overall class performance and engagement.
- Led graduate-level Data Structures & Algorithms for 40+ master's students - dynamic programming, graph theory, complexity.
- Held weekly sessions assisting 20+ students/week with problem-solving and Java assignments.
- Built and maintained Circuit, a Java backend integrated with Brightspace LMS via REST APIs, enabling anonymous peer review for 10,000+ students.
- Designed API contracts and integration patterns handling edge cases for concurrent submissions; wrote JUnit test suites.
- Built Variate, a STEM assessment platform on PostgreSQL generating per-student randomized problem sets.
- Automated provisioning with Terraform and Ansible, cutting deployment time 3× and saving 130+ hours/year across research teams.
- Deployed on-prem testbeds to simulate and monitor system behavior under threat scenarios.
- Cut infrastructure storage costs 20% through IaC tooling and disciplined configuration management.
- TA'd CS 251 Data Structures & Algorithms - developed curriculum and graded for 200+ students.
- Held 3+ weekly office hours resolving student queries and debugging assignments.