AI Engineer Intern - Skin Color & Image Analytics

TEOXANE Laboratories; Genf; On-site; CHF 20,000 - 31,000; Part time

Job summary

Teoxane Laboratories in Geneva, Switzerland, offers a 6-month internship in the IT department within the AI Lab. You will build an image processing pipeline to measure skin color from clinical photos and develop a probabilistic model with quantified uncertainty.

Responsibilities include quality gating, variance analysis, cross-system calibration, and contributing to Noesis and LightRAG initiatives. Strong Python and colorimetry interest are essential.

Qualifications

• Ongoing Master’s in Data Science, Applied Statistics, Engineering, or a related field.
• Strong Python: NumPy, pandas, scikit-learn, and image processing (OpenCV, Pillow).
• Solid applied statistics: variance components, uncertainty quantification, calibration.
• Familiarity with PyTorch; interest in colorimetry (CIELAB), LLMs and RAG is an asset.
• Curious, rigorous, autonomous; works well independently and in a cross-functional team.

Responsibilities

• Build an image processing pipeline sampling skin color at multiple facial locations with automated quality gating.
• Quantify measurement precision: variance decomposition across subject, site, imaging system, viewing angle and visit.
• Characterize and correct differences between imaging systems.
• Develop a calibrated probabilistic model estimating Fitzpatrick skin type with explicit uncertainty.
• Run robustness and fairness checks across the measured skin tone range.
• Contribute to other AI Lab initiatives and document methodology, results, and code.

Skills

Python

Image processing

Statistics

Colorimetry

Cross-functional teamwork

Education

Master's in Data Science

Tools

OpenCV

Pillow

NumPy

pandas

scikit-learn

PyTorch

Job description

Teoxane Laboratories in Geneva, Switzerland, offers a 6-month internship in the IT department within the AI Lab. You will build an image processing pipeline to measure skin color from clinical photos and develop a probabilistic model with quantified uncertainty.

Responsibilities include quality gating, variance analysis, cross-system calibration, and contributing to Noesis and LightRAG initiatives. Strong Python and colorimetry interest are essential.