Synthetic Data Engineer- Medical Imaging
Role summary
MednTech is seeking a volunteer Synthetic Data Engineer specializing in Medical Imaging to enhance their AI-powered cervical cancer screening tool. This role involves designing and implementing synthetic data generation pipelines using diffusion models or GANs to augment limited cervical image datasets and improve model robustness. Key responsibilities include generating class-balanced datasets, evaluating generative models using metrics like FID and KID, performing domain gap analysis, and collaborating with ML engineers. The ideal candidate will have experience with PyTorch, Hugging Face Diffusers or GAN frameworks, and a strong understanding of data augmentation versus synthetic generation tradeoffs. Familiarity with medical imaging challenges is preferred.
Volunteer for a startup using AI to save lives from cervical cancer! 🩺
MednTech
provides an Android AI tool that analyzes cervical images to support frontline screening decisions and improve health outcomes for women in low-resource settings.
Cervical cancer is preventable, yet it claimed 350,000 lives in 2022—94% of which occurred in low- and middle-income countries. In sub-Saharan Africa, it remains the leading cause of cancer death among women, yet screening rates are as low as 14% due to a lack of infrastructure and trained providers. While visual inspection is the WHO-recommended screening method, its interpretation is highly subjective and inconsistent, leaving frontline workers without the necessary quality assurance to make life-saving decisions at scale.
MednTech addresses this crisis through CerviScanner, an AI-based diagnostic support tool that embeds expert-level computer vision directly into the existing clinical workflow. Using a standard Android smartphone, frontline health workers can capture cervical images and receive real-time classification and confidence scores without the need for specialized labs or onsite specialists. Built on locally collected, expert-labeled data, the app aligns with national protocols and flags uncertain cases to ensure clinicians remain in control of the final diagnosis. By providing objective, AI-assisted feedback where it is needed most, CerviScanner empowers healthcare providers to deliver high-quality, accessible screening that can prevent thousands of unnecessary deaths.
Role (Volunteer, unpaid): Synthetic Data Engineer-Medical Imaging
Role Description:
We are looking for a data professional to design and implement synthetic data generation pipelines. These images will be used to augment limited cervical image datasets and improve the robustness of our cervical cancer screening model.
*Key Responsibilities*
- Develop synthetic image pipelines using diffusion models (Stable Diffusion, DreamBooth, LoRA) or GAN-based approaches
- Generate class-balanced datasets, particularly for underrepresented abnormal cases
- Design and run experiments comparing real vs synthetic vs hybrid datasets
- Evaluate synthetic data quality using FID, KID, and downstream model performance
- Implement prompt engineering and conditioning strategies for medically realistic outputs
- Perform domain gap analysis between real and synthetic data
- Collaborate with ML engineers to integrate synthetic data into training pipelines
*Required Skills*
- Experience with PyTorch and Hugging Face Diffusers or GAN frameworks
- Strong understanding of data augmentation vs synthetic generation tradeoffs
- Experience evaluating generative models (FID, distribution alignment)
- Familiarity with medical imaging challenges (preferred but not required)
*Preferred Qualifications*
- Experience with GAN based models and LoRA fine-tuning pipelines
- Prior work with imbalanced or medical datasets
- Understanding of bias in synthetic data
All roles are highly collaborative and will work closely across the MednTech AI pipeline
Experience with healthcare AI, low-resource environments, or global health applications is a strong plus
Candidates should be comfortable working in fast-paced, early-stage environments
Time Commitment: Volunteer 4-6 hours per week for a one-off project remotely 💻
If you want to make change happen, apply to volunteer with MednTech now!