- •Designed the AI backbone for CSIR's Comprehensive Geotechnical Assessment of NH 208 and 108B in Tripura, a national infrastructure project evaluating soil stability and collapse risk across 200km of strategic highway
- •Processed terabytes of satellite imagery, LiDAR point clouds, and borehole data through custom ML pipelines that replaced months of manual geological surveys with automated inference in under 48 hours
- •Built computer vision models that flagged high-risk failure zones with 94% accuracy, giving engineers actionable intelligence before a single shovel hit the ground
- •Shipped a live dashboard that CSIR teams now use to monitor geotechnical risk heatmaps across the entire corridor in real time
PythonComputer VisionGISTensorFlowData Pipeline