Generative AI Engineer

About the Role: Build generative AI solutions for future doctors. You will own the creation of high-quality medical assets—tutors, clinical cases, and exam questions—grounded in rigorous medical accuracy reviewed by clinicians.

What This Is: Starting as a stipend-based internship (approx. Rs 20k + incentives) with potential conversion to a founding engineer role offering equity. The work is high-stakes: wrong generated images teach wrong medicine, and incorrect questions mislead students.

You Will Build

• Medical Image Generation: Fine-tune diffusion models (LoRA, DreamBooth) for anatomy, histology, radiology teaching materials, ensuring strict anatomical accuracy.
• Clinical Content: Generate exam-grade vignettes and questions (NEET-PG, USMLE style), manage multimodal tutors capable of analyzing X-rays/ECGs, and build versioned datasets.

Tech Stack & Requirements

• Models: Experience with Vision-Language models (Qwen-VL, LLaVA) and Diffusion pipelines. Ability to handle fine-tuning on rented GPUs (RunPod/AWS).
• Data Engineering: Build ingestion/cleaning pipelines for de-identified medical content using Python/PyTorch.

Safety & Ethics

You will ensure every generated asset is labeled as synthetic, undergo rigorous evaluation by doctors to prevent hallucinations, and maintain fairness across demographics. Medical background is not required; curiosity about learning medicine and strong engineering foundations are essential.