Ritik Bompilwar

Ritik Bompilwar

AI Researcher & Engineer
Video Understanding Multimodal AI Agentic Harness

I'm Ritik, an AI researcher and engineer working across computer vision, multimodal AI, and applied generative systems. I recently completed my MS in Artificial Intelligence at Northeastern University, where my thesis, CrickFACT: Temporal Action Segmentation Guided Framework for Cricket Shot Recognition, focused on fine-grained video understanding in cricket under the guidance of Dr. Ehsan Elhamifar and Dr. Habib Noorbhai. My research interests are in video understanding, multimodal AI, and evaluation systems for AI-generated media. Most recently, my work includes the VQualA 2025 AIGC Challenge paper and QualiVision, which placed on the challenge leaderboard.

LinkedIn X GitHub Scholar
Career

Experience

Jan 2025 — Jun 2025

Gen AI Product Developer, Co-op

Burnes Center for Social Change
Boston, MA
  • Collaborated with Massachusetts Operational Services Division to streamline state procurement and vendor submission review
  • Applied human-centered design to map reviewer workflows, identify bottlenecks, and define product requirements
  • Architected One L, an automated legal review agent that flags terms conflicts, cites references, and generates redlines
  • Built the system as a production-grade multi-agent RAG workflow on AWS, reducing first-pass review time by 83% and enabling 6x review capacity
  • Implemented event-driven services with AWS Lambda, API Gateway WebSockets, CloudWatch, Titan embeddings, and OpenSearch vector search
  • Developed React and Amazon Cognito workflows for role-based access, auditability, and step-by-step reviewer traceability
  • Integrated RAGAS-based LLM evaluation and production KPIs into Assistive Buyer's Engine, then deployed it to production
Jan 2023 — Apr 2023

Jr. AI Developer Intern

Openfabric AI
Remote
  • Developed AI applications for Openfabric's distributed computing platform and marketplace
  • Adapted open-source ML projects into user-facing apps with custom UIs using the Openfabric toolkit
  • Optimized model performance and deployment workflows for production-ready marketplace distribution
  • Containerized and deployed Deep Fake AI Generator, AI Cryptobot, AI Face ID Recognition, and AI 3D Human Rotation Generator using Docker
May 2022 — Aug 2022

MITACS Globalink Research Intern

Dalhousie University
Truro, Canada
Advisor: Travis Esau (Professor and Director, Atlantic Institute for Digital Agriculture)
  • Engineered a machine vision system for weed detection in wild blueberry fields
  • Built localized spraying workflows to help farmers reduce weedicides and operational costs
  • Created semi-supervised annotation workflows to reduce labeling effort for large-scale weed detection data
  • Trained and benchmarked YOLOv5, YOLOv4, and EfficientDet Lite, achieving 85.8% mAP with YOLOv5 n6
  • Optimized model size by 10x through quantization for constrained edge deployment
  • Built live RGB camera pipelines with FLIR Firefly, Blackfly, and OAK-D for real-time field inference
  • Deployed optimized ONNX, TensorRT, and TF Lite models on Jetson Nano, achieving 17 FPS weed detection
May 2021 — Apr 2022

Computer Vision Intern

Big Vision LLC
Remote
Mentor: Dr. Satya Mallick (CEO, OpenCV & Big Vision)
  • Designed computer vision and deep learning projects for OpenCV University's applied AI curriculum
  • Built a radiology AI model to classify chest X-rays into COVID-19, pneumonia, and normal cases
  • Developed a small ship detection model for aerial imagery using TensorFlow Object Detection API, achieving 82.5% mAP
  • Built a fashion apparel classification model across five classes, achieving 99% accuracy
  • Authored three LearnOpenCV articles on TensorFlow Lite model optimization for edge deployment
Academic

Education

Jan 2024 — Apr 2026

Master of Science in Artificial Intelligence (Research)

Northeastern University, Khoury College of Computer Sciences
Boston, USA
GPA: 3.875 / 4.0
Aug 2019 — May 2023

Bachelor of Technology

International Institute of Information Technology (IIIT), Naya Raipur
Naya Raipur, India
GPA: 8.77 / 10