I build intelligent systems that feel seamless, reliable, scalable, and thoughtfully engineered from idea to deployment.
I’m an AI Engineer based in Nepal, building intelligent systems that turn ideas and data into real-world solutions. I enjoy working beyond the model — understanding data, designing effective solutions, building APIs and pipelines, and bringing AI into production.
Outside of work, I explore machine learning, computer vision, NLP, and generative AI through projects and experiments. There’s always something new to understand, improve, or turn into a working system.
Right now, I’m focused on building AI that is not only intelligent, but reliable, practical, and built to be used.
Becoming better than yesterday.
What I reach for, and what I'm still learning.
Things I've designed and built end to end.
AI-powered adaptive exercise coaching system for real-time pose analysis and personalized guidance. Implemented joint angle estimation with EMA stabilization, cosine similarity-based recommendations, and a dynamic engine adjusting to user demographics and posture patterns.
Python
React
DjangoMediaPipe
PostgreSQL
AI-driven pipeline generating PBR texture maps (depth, normal, roughness, ambient occlusion) from a 2D image. Implemented a hybrid Vision Transformer architecture using DINOv2 encoder and custom feature-fusion modules with RAFT for high-resolution single-step inference.
Python
PyTorchDINOv2Vision TransformerWhere I've worked and what I've accomplished.
