Jan – Jul 2026
Clinical AI R&D Intern
Harvard Medical School · Boston
Open to new roles
I'm Varun Teja Kalakoti, a computer vision engineer in Boston. I build models that see, for medical AI, industrial automation and autonomous systems.
01 — Selected work
Compared five deep learning models for carotid artery segmentation, then rebuilt coronary arteries in 3D from two X-ray views.
0.92 Dice score, carotid segmentation
View case studyAnswers questions across 304 PDF pages and 501 images, and cites the sources behind each answer.
~40% faster image queries (2.2s → 1.3s)
View case studyDetecting and tracking six coral families in underwater video, so autonomous underwater vehicles can monitor reefs.
97.3% mAP@0.5 · published in Current Science
View case studyClassifying skin lesions as benign or malignant with an EfficientNet that uses 92.6% fewer trainable parameters.
93.7% test accuracy · published in IEEE
View case studyDefect detection running live on NVIDIA Jetson: gas-leak bubbles, cracked gear teeth and car-panel defects.
Real-time on Jetson Nano and Orin, linked to factory PLCs
View case studySegmenting arteries at Harvard Medical School. Inspecting parts on factory floors. Tracking coral for underwater robots. Now I'm bringing language and vision together with retrieval and agents.
Jan – Jul 2026
Clinical AI R&D Intern
Harvard Medical School · Boston
Mar – Jun 2024
Computer Vision Engineer Intern
Eternal Robotics · Hyderabad
Feb – Jul 2023
Research Intern
AppsTek Corp · Hyderabad
Expected Dec 2026
MS, Information Systems
Northeastern University
July 2024
B.Tech, Data Science & AI
Woxsen University
Exploring
Language × vision
LLMs, retrieval, agents
03 — Toolkit
YOLOv8, nnU-Net, SegMambaV2, EfficientNet, ByteTrack, segmentation, tracking
Python, C, C++, SQL, PyTorch, TensorFlow, Keras, OpenCV, scikit-learn, Transformers
Docker, Kubernetes, TensorRT, DeepStream, Jetson, FastAPI, AWS, MLflow, CI/CD
RAG, LangGraph, pgvector, CLIP, Ollama, RAGAS, agents
04 — Contact