Hi, I'm Himanshu Ranjan
MS Computer Science @ UMass Amherst
Former Machine Learning Engineer with experience building real-world ML systems
I’m interested in developing machine learning systems that combine perception, reasoning, and decision-making, and can operate reliably in complex environments.
- Machine Learning and Deep Learning
- Computer Vision and 3D Scene Understanding
- Reinforcement Learning and Sequential Decision-Making
- Large Language Models and Multimodal Learning
- Probabilistic Modeling and Reasoning under Uncertainty
- Scalable ML Systems and Model Optimization
- Built and improved learning-based systems for perception and structured prediction
- Developed models for depth estimation, representation learning, and 3D understanding
- Applied probabilistic approaches for state estimation and decision-making under uncertainty
- Designed and optimized end-to-end pipelines for training, evaluation, and deployment
- Worked on multimodal systems combining vision and language for information extraction and reasoning
Languages
Python, C++
Machine Learning & AI
Deep Learning, Computer Vision, 3D Vision, Reinforcement Learning, Generative AI,
Large Language Models (LLMs), Transformers, Diffusion Models, Retrieval-Augmented Generation (RAG),
Representation Learning, Model Optimization
Frameworks & Libraries
PyTorch, TensorFlow, Hugging Face, OpenCV, Open3D
Systems & Tools
Docker, Git, Linux, Azure, ONNX, Triton Inference Server
Development & Interfaces
Flask, FastAPI, Streamlit