About
I'm a Senior Software Engineer specializing in 3D Computer Vision and photogrammetry. I build systems that bridge the gap between the physical and digital worlds through advanced 3D processing and web technologies.
Bio
I live in Chennai, India, where I work as a Senior Software Engineer in 3D Computer Vision at Eicon Vision. I gratuated as a Mechanical Engineer and then moved to the software world two years later. My journey led me from traditional engineering into the fascinating world of 3D computer vision and web development.
I spend my days working on computer vision projects, developing 3D reconstruction pipelines and creating interactive 3D web applications. I've built systems that handle high-volume data processing and developed solutions that have saved clients millions of dollars annually.
My expertise spans Python, Rust, JavaScript, 3D processing tools like Metashape and Blender, and modern web frameworks. I'm passionate about automating complex 3D workflows and making advanced computer vision accessible through intuitive web interfaces.
Projects
3D Processing Automation Pipeline
Built an end-to-end industrial 3D scan processing system with REST API integrations. Designed and maintained 3D photogrammetry pipelines improving scan success rates from 50% to 90%. Automated 3D measurement workflows and integrated CI/CD pipelines, increasing throughput by 40%.
Technologies: Python, OpenCV, Metashape, PyTorch, Node.js
3D Wound Care Application
A cross-platform application for 3D wound analysis and measurement. Built web-based 3D viewers using three.js and integrated 3D processing workflows for automated wound size estimation and visualization. Handled high-volume 3GB video input and 2GB model data processing.
Technologies: Three.js, Python, Node.js, Flutter, FastAPI, PyTorch, Metashape
Blender Photogrammetry Add-On
Automated point-cloud preprocessing, mesh decimation, and texture projections for 3D models. Delivered a GUI-based batch processing system that reduced manual processing time by 60%. Created automated mesh cleanup and photogrammetry preparation workflows.
Technologies: Python, Blender API, PyTorch, SAM
Image Retrieval System for Artworks
Built a large-scale image retrieval system for artworks and artifacts with a database of 1.5 million images. Developed image analysis pipelines for art authentication and 3D artifact reconstruction. Created interactive three.js-based web modules for real-time 3D model viewing and annotation.
Technologies: Python, Three.js, Computer Vision, Image Processing
Blogs
Building visual search for inventory management with Qdrant
February 25, 2025
How I'm solving the "where the hell is that thing" problem with vector databases, and why traditional inventory systems are fundamentally broken. Built a system where you can take a photo of anything and find it in your inventory in under 500ms.
Greedy Algorithms, and When "Good Enough" is Perfect
January 15, 2024
A practical exploration of using greedy algorithms for mechanical part dimensioning. Sometimes choosing the best local option at each step gives you a solution that's either optimal or good enough to be indistinguishable from optimal.
Get in Touch
I'm always open to discussing new opportunities, collaborations, or interesting projects. Feel free to reach out through any of these channels.