Results-driven Computer Vision & AI Engineer with 3+ years of experience in the development and optimization of computer vision with deep learning models and large-scale AI systems (HPC). Expertise in parallel computing, photogrammetry and computational geometry, and real-time vision applications for autonomous systems. Passionate about parallel computing, GPU acceleration (CUDA, TensorRT) and ADAS. Proven success in improving model efficiency, scalability, and deployment across NVIDIA Jetson, AWS, and embedded AI environments.


Vehicle, Lane and Pedestrian Detection (ADAS)

πŸš— Vehicle RoadSense – Vehicle, Lane and Pedestrian Detection

Real-time ADAS system integrating classical image processing with YOLOv8 for robust multi-object detection and lane estimation.

Python OpenCV YOLOv8
Learn More β†’
Pipeline Parallelism for Large Language Models (GPipe)

🧠 Pipeline Parallelism for LLMs (GPipe)

Experimental GPipe-style pipeline parallelism for GPT-like models with substantial throughput and memory improvements.

PyTorch CUDA Distributed
Learn More β†’
Automated Body Measurement Estimation using YOLOv5 and Pose Estimation

πŸ“ Automated Body Measurement Estimation (YOLOv5 + Pose)

GPU-accelerated measurement system combining YOLOv5 and pose estimation with multi-GPU HPC deployment.

YOLOv5 OpenPose CUDA
Learn More β†’
3D Point Cloud Segmentation via 2D Segmentation Fusion

🧩 3D Point Cloud Segmentation using 2D Segmentation

Hybrid OneFormer-based 2D segmentation fused into 3D with COLMAP-style voting for efficient point cloud labeling.

OneFormer Point Cloud COLMAP
Learn More β†’
View All Projects β†’

πŸ› οΈ Core Technical Skills

πŸ€– Deep Learning & AI

  • β€’ PyTorch, TensorFlow, Keras
  • β€’ CNNs, GANs, Transformers, YOLO
  • β€’ NeRF, Transfer Learning
  • β€’ Multi-GPU Training, Pipeline Parallelism

πŸ‘οΈ Computer Vision

  • β€’ OpenCV, Open3D, SLAM
  • β€’ 3D Reconstruction, Photogrammetry
  • β€’ Object Detection, Segmentation
  • β€’ Point Cloud Processing, Depth Estimation

⚑ High Performance Computing

  • β€’ CUDA, TensorRT, GPU Acceleration
  • β€’ Distributed Computing, HPC Clusters
  • β€’ Docker, Kubernetes
  • β€’ NVIDIA Jetson, Edge Deployment

πŸ’» Programming & Development

  • β€’ Python (Expert), C/C++, CUDA
  • β€’ Git, GitLab, Shell Scripting
  • β€’ SQL, LaTeX, Jupyter
  • β€’ LLM, RAG, Multi-modal AI

πŸ”§ Specialized Hardware

  • β€’ iPhone LiDAR, Intel RealSense
  • β€’ RGB-D Cameras, Multi-Camera Systems
  • β€’ Sensor Fusion, IMU Integration
  • β€’ Raspberry Pi NPU, ROS

🎯 Domain Expertise

  • β€’ Autonomous Driving (ADAS)
  • β€’ Medical AI & Surgical Assessment
  • β€’ Underwater Image Processing
  • β€’ AR/VR, 3D Content Generation
View Detailed Skills β†’

πŸŽ“ Education & Background

πŸŽ“

Master of Science in Computer Science

Case Western Reserve University

2023 - 2025

Specialization in Computer Vision and LLM Parallelization

  • Key Projects: Advanced 3D Object Detection, GPipe LLM Parallelization, ADAS Lane & Vehicle Detection
  • Relevant Coursework: Deep Learning, Advanced Computer Vision, Robotics
πŸŽ“

Bachelor of Engineering in Computer Science & Engineering

KLE Technological University

2017 - 2021

  • Member of Center of Excellence in Visual Intelligence (CEVI)

πŸ† Professional Experience Highlights

AI Engineer Intern Vision LLM

IdeaBoxAI β€’ Jun 2025 - Jul 2025

Prototyped agentic AI system for automated insurance claim liability analysis, implementing multi-modal evidence processing with LLM orchestration and RAG for evidence-based recommendations.

3D AI Lead Engineer

GPU.net (formerly brahmGAN) β€’ Jun 2023 - Feb 2024

Architected parallelized deep learning pipelines for text-to-3D, image-to-3D, and video-to-3D model generation. Designed decentralized GPU computing framework, reducing cloud costs by 30%.

Software Engineer Computer Vision

Tooliqa Innovations LLP β€’ Mar 2021 - Apr 2023

Developed 360 Video-to-3D Model Pipeline using hybrid Deep Learning + Classical Computer Vision. Achieved 90% background noise reduction and 25% trajectory accuracy improvement through sensor fusion optimization.

Research Project Intern

KLE Technological University β€’ Aug 2020 - Feb 2021

Designed UIE-Net GAN-based model for underwater image enhancement, achieving 60-75% improvement in image visibility. Research accepted at GC-RDCT-21 conference.

Project Trainee Intern

Indian Institute of Technology Delhi β€’ Jun 2019 - Jul 2019

Research-based medical imaging collaboration with AIIMS Delhi. Built automated AI-driven surgical micro-suturing assessment tool for neurosurgical training applications.

Key Academic Achievements

Case Western Reserve University β€’ 2023 - 2025

42% speedup in LLM training throughput through pipeline parallelism. Advanced 3D point cloud segmentation using hybrid 2D-3D approaches. Real-time ADAS system with YOLOv8 integration.


πŸ“¬ Let’s Connect

Ready to Collaborate?

I’m always excited to discuss new opportunities, innovative projects, or potential collaborations in computer vision and AI.

Response time: Usually within 24 hours | Open to remote opportunities worldwide


πŸ”— Quick Navigation


🌟 This portfolio showcases my passion for computer vision and AI β€’ Built with Quartz 4

Last updated: β€’ Made with ❀️ for innovation