Muhammad Saad Jahangir

Lysaker, Norway

• Designed and implemented scalable backend services for high-impact use cases like predictive maintenance, integrating physics-based models, calculation engines, simulations, and machine learning.
• Integrated AI into operator and analyst workflows, improving decision-making and planning processes.
• Participated in customer workshops to understand use cases and shape product design.
• Mentored AI summer students and supported teams in AI-focused hackathons.

Tech stack: TypeScript, Python, Node.js, Mongo, Timescale, Qdrant, FastAPI, Express, Angular

Lysaker, Norway

• Developed a document parser using OCR and deep learning based symbol detection to extract equipment and sensor information from piping and instrumentation diagrams.

Tech stack: Python, PyTorch, FastAPI

Ålesund, Norway

• Plastopol: Identified sweet spots and predicted trends for marine litter accumulation along the coast of Møre og Romsdal.
• Teaching Assistant for IMAA2150 - Mathematical Methods 3 for computer engineers.
• Towed ROV: Trained and deployed machine learning models for underwater trash detection.

Built a camera and sensor based IoT system running state-of-the-art models on a Jetson Nano, assessing driving behaviour through acceleration, deceleration, turning, lane departure, tailgating, and driver drowsiness/attentiveness detection.

Built a Unity 3D city simulator for visualizing spatiotemporal urban data, including landmark visibility, light intensity, shadow heatmaps, and sky exposure, with runtime model updates for what-if scenarios.

Developed a visualization platform as a digital twin of Ålesund City for landmark visibility, shadow variation, and bus stop accessibility analysis.

Norwegian University of Science and Technology, Ålesund, Norway

Specialization in Machine Learning and Computer Vision. Thesis: Targetless Multi-LiDAR Calibration for Autonomous Trucks.

University of Engineering and Technology Lahore, Lahore, Pakistan