x .png

Hi there! I’m Himanshu, 21

About Me

Fusing raw grit with sharp skills to build optimized ML/DL pipelines that actually deliver. Relentlessly grinding daily to master new models. Just trying my way out through every niche.

Get in touch

<aside> <img src="/icons/mail_gray.svg" alt="/icons/mail_gray.svg" width="40px" /> hs262764@gmail.com

</aside>

<aside> <img src="notion://custom_emoji/cd18771a-518b-4774-a7eb-387723b9cb9d/2a88244c-f5d6-8087-a640-007a61acb88c" alt="notion://custom_emoji/cd18771a-518b-4774-a7eb-387723b9cb9d/2a88244c-f5d6-8087-a640-007a61acb88c" width="40px" /> Himanshu

</aside>

<aside> <img src="/icons/passport_gray.svg" alt="/icons/passport_gray.svg" width="40px" /> Website

</aside>

</aside>

<aside> <img src="notion://custom_emoji/cd18771a-518b-4774-a7eb-387723b9cb9d/2a88244c-f5d6-80e7-9fc0-007abb6531c4" alt="notion://custom_emoji/cd18771a-518b-4774-a7eb-387723b9cb9d/2a88244c-f5d6-80e7-9fc0-007abb6531c4" width="40px" /> Github

</aside>

<aside> <img src="/icons/phone-call_gray.svg" alt="/icons/phone-call_gray.svg" width="40px" /> Phone

</aside>

Work Experience

<aside> Ai engineer intern

mytender (July - Sep 2025)

Improved RAG performance and retrieval accuracy through adaptive optimization.
Built a scalable front-end enabling smoother multi-module interactions.
Enhanced agent workflows and orchestration quality across systems.
Streamlined data pipelines for faster, consistent, and reliable processing
Tools: LangChain (Python), Next.js, ChromaDB, BM25. </aside>

<aside> AI/ML intern TEXMiN (May - July 2025)

Built YOLO-based traffic detection for real-time tracking on limited hardware.
Trained custom YOLO model for mining equipment with efficient performance.
Deployed optimized models on Jetson Nano using TensorRT for smooth inference.
Improved YOLO runtime on Raspberry Pi through pruning and NCNN optimization.
Tools: PyTorch, Ultralytics, TensorRT, NCNN, OpenCV, ONNX, Roboflow, C++, Python. </aside>

<aside> Research Intern IIT Bhilai (May - August 2024)

Built advanced image forgery detection models with TensorFlow for high-accuracy classification.
Optimized algorithms and workflows to reduce processing time and improve execution.
Managed large image datasets with efficient preprocessing and annotation pipelines.
Tools: PyTorch, TensorFlow, and OpenCV for model development and training. </aside>