Hi, I'm Samunder Singh

Data Scientist

Building intelligent systems and scalable machine learning solutions with expertise in computer vision and deep learning. I'm also an active open source contributor, passionate about building tools that empower the developer community. Transforming data into actionable insights while contributing to meaningful open source projects.

Samunder Singh

About Me

I am a Data Scientist with a strong background in computer vision, deep learning, and scalable ML systems. Over the past year, I’ve developed and integrated advanced models—including YOLO, DETR, and PaddleOCR—into production-grade inspection pipelines. My work spans from designing self-supervised and unsupervised learning algorithms (e.g., CLIP-based anomaly detection, auto-labeling) to scaling infrastructure across multi-GPU and multi-host environments.

I bring hands-on experience with backend frameworks like Django and FastAPI, and technologies such as MongoDB, Docker, and AWS, enabling seamless end-to-end deployment of AI solutions.

In addition to my industry work, I'm an active contributor to the open-source ML community, with 80+ pull requests merged into ivy, and contributions to keras and crossgl.

Work Experience

Lincode.ai

Bengaluru, India (Hybrid)
May 2024 - Present
Associate Data Scientist (May 2024 - Present)
  • Designed and deployed YOLO-v8 and DETR object-detection models for real-time visual inspection in production.
  • Built automated image-segmentation and OCR pipelines with PaddleOCR, improving text-extraction accuracy.
  • Fast-tracked to full-time; architected a CLIP-driven auto-labeling and anomaly-detection workflow, reducing manual labeling effort.
  • Optimized computer-vision pipelines for distributed multi-GPU and multi-node clusters, lowering inference latency.
  • Refactored Django + MongoDB back-end into a modular microservice architecture, accelerating deployment cycles.

CrossGL

Remote
Jun 2024 - May 2025
Open Source Developer
  • Contributed to CrossTL: implemented a universal shader intermediate representation (IR) CrossGL that enables translation between any shader language and the CrossGL IR, and back.
  • Enhanced the Metal, OpenGL, DirectX, and Vulkan translators, adding 25,000 lines of code across 64 commits.
  • Helped develop Dyson—a dynamic compute router that dispatches workloads to any available device.

Unify.AI

Remote
Jun 2023 - Present
Open Source Developer
  • Contributed 80+ PRs across JAX, PyTorch, TensorFlow, and PaddlePaddle
  • Contributed in building LLM benchmarking chatbot with leaderboard and voting
  • Contributed in intergrating the backend and ivy unfiy api with torch , jax , tensorflow , paddlepaddle and numpy

My Open Source Contributions

Featured Projects

SLX

Slang PyTorch Python CUDA

A high-performance deep learning framework built on top of PyTorch, designed for efficient tensor operations with support for custom shader-based kernels through Slang.

Deep Learning View Project

Shruti-Drishti

Python MediaPipe TensorFlow Flask

App for real-time ISL recognition using Mediapipe, CNN, and LSTM with 90%+ accuracy. Won award at Hasgeek hackathon for developing real-time Indian Sign Language detection using Mediapipe Holistic.

Computer Vision View Project

Edulens

FastAPI Celery Google OAuth vetorDB

Edulens is a FastAPI video analysis platform that uses ML for figure detection, OCR, and speech-to-text, with Google OAuth for secure access and Vectara integration for enhanced language processing.

Computer Vision/NLP/LLM View Project

LLM Playground

Docker GCP Streamlit LLMs

Chatbot Arena: Benchmarking LLMs in the Wild. A chat interface where users enter prompts and simultaneously see outputs from two chosen LLMs for comparison.

NLP / LLM View Project

Sign Language Detection

Python MediaPipe TensorFlow Flask

Translated real-time sign gestures into text using holistic landmark tracking with MediaPipe and TensorFlow.

Computer Vision View Project

Travel Chatbot with RAG

LangChain OpenAI Streamlit RAG

Developed an LLM-based chatbot fetching weather and search information in real-time using LangChain and OpenAI.

NLP / LLM View Project

YOLOv7 Custom Training

YOLOv7 PyTorch W&B OpenCV

Trained object detection model on custom aquarium dataset with comprehensive visual logging and monitoring.

Computer Vision View Project

Technical Skills

Languages

Python
SQL
C++
CUDA

ML Frameworks

PyTorch
TensorFlow
JAX
PaddlePaddle
NumPy

Computer Vision

YOLOv5/v7
CLIP
MediaPipe
OpenCV
PaddleOCR

Infrastructure

Docker
GitHub Actions
GCP
AWS
Terraform

Web & Tools

Django FastAPI Flask Streamlit Redis MongoDB PostgreSQL

Certifications

Latest Articles

Recommendations

DL

Daniel Lenton

May 6, 2024

CEO at Unify | YC W23

Samunder has done great work during his time as a top contributor working on Ivy. His enthusiasm was infectious to those around him, and he learned a lot during the role, picking up new skills very quickly, with a drive to get things done. I would recommend him for any role involving Machine Learning infrastructure, and especially those using a lot of Python!

Get In Touch

Have a project in mind or want to chat? Feel free to reach out!

Send me an email

Or find me on social media: