About the Role

We are looking for a Medical AI Evaluation Intern to benchmark large medical vision-language models on real-world clinical imaging data. You will work with a dataset of intraoral endoscopic images, run inference using state-of-the-art medical multimodal models, and systematically evaluate their performance on disease detection and classification tasks.

This hands-on position is ideal for students or early-career researchers interested in the intersection of medical imaging, large language models, and clinical AI evaluation. You will gain direct exposure to the full evaluation pipeline — from data preparation and prompt engineering to metric design and analysis.

Responsibilities

- Curate, clean, and annotate a dataset of intraoral endoscopic images following clinical labeling guidelines.

- Design and execute evaluation protocols for medical vision-language models, including ground-truth annotation, inference, and performance analysis.

- Perform prompt engineering and workflow design to optimize model outputs for disease identification and classification.

- Run systematic experiments to measure accuracy, recall, precision, and failure modes across different anatomical regions and pathological conditions.

- Document findings in clear, reproducible reports and present results to the team.

Minimum Qualifications:

- Currently pursuing or recently completed a Bachelor's or Master's degree in Computer Science, Biomedical Engineering, Electrical Engineering, Data Science, or a related field.

- Proficiency in Python.

- Basic understanding of computer vision or medical image analysis concepts (classification, detection, segmentation, evaluation metrics).

- Comfortable working with image datasets: loading, preprocessing, visualization, and annotation.

- Strong attention to detail and ability to systematically document experimental procedures and results.

Preferred Qualifications:

- Experience with large language models or vision-language models (GPT, Gemini, LLaVA, Med-PaLM, or similar), including prompt engineering and API usage.

- Exposure to medical imaging or healthcare AI projects.

- Familiarity with at least one deep learning framework.

- Experience with experiment tracking tools (Weights & Biases, MLflow, or similar).

- Ability to write lightweight Python scripts to automate data processing and evaluation workflows.

- Basic understanding of statistical analysis and experiment design.

- Mandarin Speaking is a plus

What You'll Gain

- Hands-on experience evaluating cutting-edge medical AI models on real clinical data.

- Deep understanding of how large multimodal models perform in domain-specific medical tasks.

- Mentorship from an interdisciplinary team spanning AI, software, and clinical applications.

- Practical training in industry-grade medical AI product evaluation methodology.

职位类型：兼职, 实习生, 合同工

薪资： $20.00至$25.00（每小时 ）

Work Location: Remote