DeepTek Tech Stack

AI decision-support system for radiologists using deep learning on medical imaging

Hospitals and Health Care New York 51–200 employees Founded 2017 Privately Held

DeepTek builds deep learning models for radiology, focusing on model optimization, cross-site validation, and bias assessment across diverse imaging devices and populations. The tech stack is standard for medical AI (Python, PyTorch, Keras, OpenCV) with DICOM/PACS integration for clinical workflows. Hiring is minimal and concentrated in India, with only senior-level data and engineering roles open—consistent with a mature R&D operation rather than rapid scaling.

Tech Stack 13 technologies

Core StackPython NumPy pandas scikit-learn Docker PyTorch OpenCV Pillow ONNX Keras Vision Transformer DICOM PACS

What DeepTek Is Building

◆Challenges

Bias detection
Cross-site validation
Robustness across diverse populations and imaging devices
Data quality and clinical validity

▲Active Projects

Deep learning models for medical image analysis
Model optimization for efficient inference
Advanced medical imaging solutions
Cross-site validation and domain generalization
Production pipeline integration
Error analysis and bias assessment for medical imaging
Domain adaptation and robustness strategies
Data storage, preprocessing and annotation strategies for medical imaging
Ai model integration into production workflows
Radiology decision-support system

Hiring Activity

Minimal6 roles · 0 in 30d

Department

Data

Engineering

Seniority

Senior

Company intelligence

Find more companies like DeepTek by tech stack, pain points and active projects

Get started free

About DeepTek

DeepTek develops AI decision-support systems for radiologists, helping reduce diagnostic workload and accelerate diagnosis cycles. Founded in 2017, the company operates as a 51–200-person team based in New York with significant operations and hospital partnerships in India. The product integrates with existing clinical infrastructure (PACS, RIS, DICOM standards) and centers on deep learning models for medical image analysis. Core challenges center on robustness—ensuring models generalize across different imaging devices, patient populations, and clinical sites—and on bias detection to meet clinical validation requirements.

HeadquartersNew York

Company Size51–200 employees

Founded2017

Hiring MarketsIndia

Frequently Asked Questions

What AI frameworks does DeepTek use?

PyTorch, Keras, and scikit-learn form the core ML stack. Vision Transformer architectures support image classification tasks. ONNX is used for model standardization and cross-platform deployment.

What is DeepTek working on?

Deep learning model optimization for medical imaging, cross-site validation, bias and error analysis, domain adaptation for robustness across diverse imaging devices, and integration into production radiology workflows.