Data Scientist Placement Student

Most of what we build is back-office software for property managers, including lease tracking, work orders and tenant payments. Founded in 2013, the business writes the billing, metering and customer portal software that small water and electric utilities run on.

The internship

We are building text classification and search features using LLMs and classic ML. The intern will help with experiments, evaluation and data labelling. Fully remote. You can work from anywhere in France. Normal business hours, Monday to Friday, with some flexibility. Please make sure you have a quiet place to work and a solid internet connection.

Responsibilities

• Package a trained model behind a simple FastAPI endpoint for testing.
• Fine-tune small transformer models with Hugging Face for classification tasks.
• Clean and prepare datasets in Python using pandas and NumPy.
• Run exploratory analysis and write up what you find in a Jupyter notebook.
• Train baseline models with scikit-learn and compare them against simple rules.
• Write prompts and evaluation sets for LLM features, and score the outputs.
• Pull data from BigQuery or PostgreSQL with SQL queries.

Who can apply

• Eligible to work in France.
• Currently enrolled in a bachelor's or master's degree, graduating between December 2027 and June 2028.
• Based in France for the duration of the internship.
• Available for the full internship period.
• Laptop with 8 GB RAM or more and a stable internet connection (no GPU needed, we provide cloud compute)
• Coursework or projects in machine learning, such as regression, classification and clustering.
• Pursuing a degree in Computer Science, Statistics, Mathematics, Data Science or a related field, graduating in between December 2027 and June 2028.
• Strong Python, including pandas, NumPy and Matplotlib.

Bonus points

• Understanding of A/B testing.
• Kaggle competitions or a published student research paper.
• Exposure to SageMaker, Vertex AI or Databricks notebooks.

What you'll learn

• How to judge whether a model is worth deploying, not just whether its accuracy is high.
• You will take one model from raw data to an evaluated prototype.
• Weekly mentorship from a senior data scientist.

Duration: 4 months

Start date: Jan 2027

Stipend: €1,500 - €1,550 per month

Perks

• Paid public holidays.
• Certificate of completion.
• Mentorship.

Duration: 4 months. Start: Jan 2027.

Demo job created on 11jobs for testing and demonstrations. It is not an actual opening at AG Software SAS.