Open to opportunities

SHUBHAM KUMAR

@shubhamkumar29

Message

Senior AI scientist specializing in medical imaging, deep learning, and LLM applications.

India

Message

What I'm looking for

I seek roles building production medical‑imaging AI and LLM systems with clinical impact, strong engineering practices, and opportunities to lead model deployment and validation.

I am a senior AI scientist focused on medical imaging and applied deep learning, with experience building production pipelines and clinical models. I design segmentation, detection, and tracking systems for CT and MRI using state‑of‑the‑art methods.

My work includes building low‑resolution whole heart segmentation pipelines, automated vessel landmark and centerline extraction, and MRI view classification using ResNet and self‑supervised embeddings. I have developed MedSAM integrations, RetinaNet detectors with high sensitivity, and nodule‑tracking algorithms for longitudinal chest CT analysis.

I have experience developing LLM‑driven reporting systems, multi‑agent chatbot interfaces, and ComfyUI object placement workflows, and I deploy models at scale using Docker, Kubernetes, and AWS ECS. My contributions include a granted US patent for automated LVO detection and publications and presentations at major conferences.

I prioritize robust, generalizable solutions that are deployable in clinical settings and emphasize reproducibility, efficiency, and measurable clinical performance improvements.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

ConcertAI

Apr 2025 - Present (9 months)

Built a low-resolution whole-heart segmentation and automated vessel landmark/centerline extraction pipeline for cardiac CT, and developed cardiac MRI view classification models to improve multi-dataset generalization.

3D NnUNet TotalSegmentator Python PyTorch Image Processing ResNet Self Supervised Learning Docker

Applied AI Engineer

Affogato.ai

Dec 2024 - Mar 2025 (3 months)

Developed a multi-LLM agent chatbot integrating RenderNet API and orchestrated multi-agent workflows using MCP and LangGraph; implemented object placement workflows in ComfyUI with advanced inpainting for high-fidelity results.

LangGraph ComfyUI Prompt Engineering Python API Integration

AI Scientist

Qure.ai

Jun 2021 - Nov 2024 (3 years 5 months)

Led development and deployment of medical imaging AI including MedSAM-based nodule segmentation, MAE pretraining with ViTs, RetinaNet detection (92% sensitivity), automated nodule tracking, and an LVO detection solution (90% sensitivity, 92% specificity).

Medical Imaging PyTorch MAE RetinaNet Docker Kubernetes

Summer Intern

Samsung R&D Institute Delhi

May 2020 - Jun 2020 (1 month)

Developed a Flask web application integrating FaceNet facial recognition, YOLO object detection, and a custom landmark recognition model to manage user data and search functionality.

Flask FaceNet YOLO Python web app development REST APIs Docker

Summer Research Intern

Hanyang University

May 2019 - Jul 2019 (2 months)

Implemented a proof-of-concept simulation measuring data transfer times between NVDLA sub-units and GDDR6 SDRAM to analyze accelerator memory transfer characteristics.

GDDR6 Embedded Systems C C++Research CUDA