I'm actively involved in the AI team. I've had the chance to work on a variety of different domains. -Working with state-of-the-art computer vision and temporal models to boost safety in the fleet management industry. This includes employing ML and Vision based approaches for, but not limited to, collision detection, tailgating, stop sign violation and unsafe lane changes.-Spearheaded the end-to-end development of the Forward Collision Warning (FCW) system—a pure vision-based collision warning solution—deployed on over 150,000 vehicles, improved precision from less than 10% to over 90%.-Established system workflow, collaborating with the embedded systems team to deploy ML algorithms on edge devices for real-time processing.-Enhanced system precision by conducting root cause analysis of false positives and developing multiple ML temporal filters, significantly reducing false positives, including prototyping and developing a Kalman Filter based approach to track bboxes in 2D space.-Worked on Generative AI projects utilizing LLMs for data interpretation and insights extraction, including creating an interface for real-time dynamic knowledge base queries using natural language, that would later go on to win a company-wide internal hackathon.

At DataInsight Research lab, worked on novel Computer Vision and NLP problems.Key project was low resource bilingual handwriting digitization which we published in our work, "Adapting multilingual vision language transformers for low-resource Urdu Optical Character Recognition (OCR)". Interactive demo can be found here: https://huggingface.co/spaces/musadac/VilanOCR-Urdu-English-Chinese

Programming Fundamentals (C++)Programming Fundamentals (Python)Fundamentals of Business AnalyticsDatabase Systems

At DataInsight Research Lab, worked on Computer Vision problems related to low-resource languages.

Associate Intern at Corporate Affairs Department.