sanatan kafle

I’m a Machine Learning Engineer focused on the complete ML lifecycle, from data engineering to production deployment. I bring strong MLOps discipline and cloud infrastructure know-how, combining Deep Learning with Computer Vision and NLP to ship scalable solutions.

At ShrigSolutions (17th Sep 2024 - Present), I designed and trained vision models for object detection, segmentation, classification, and depth estimation. I also engineered 3D length estimation using LiDAR with AR, optimized models for low-latency edge inference on mobile, and streamlined the end-to-end ML workflow by building automated AWS-based data ingestion, training, and deployment pipelines with MLflow and Grafana Cloud.

I deployed ML architectures on AWS and developed backend REST APIs using FastAPI to serve model inference to production through microservices. In my earlier role (Junior, 7th Jun 2024 – 16th Sep 2024), I built YOLO-based computer vision systems for object detection, pose estimation, and instance segmentation, plus monocular depth estimation.

I’ve also worked on research and development like video inpainting, duplicate image detections, and Vision Transformers, and I enjoy mentorship—training and mentoring interns and trainees on machine learning concepts. My goal is to keep building robust, production-grade models with practical engineering at every step.