Job Description

What you'll do:

You will work embedded in the POI team on concrete research and engineering tasks, with guidance from senior engineers/scientists. You will:
Explore and experiment with ML/AI approaches to solve POI-domain problems such as entity matching, address parsing, data quality assessment, or coverage analysis
Implement and evaluate models and algorithmic solutions on real-world, large-scale geospatial datasets
Design and run experiments, analyze results, and translate findings into clear insights, recommendations and implementation
Be part of the development of data pipelines and tooling that support model training, evaluation, and analysis
Collaborate with Applied Scientists, Engineers, and Product stakeholders to understand requirements and integrate your work into the broader team workflow
Document experiments, methodologies, and results clearly to support knowledge sharing within the team

What you'll need:

Currently enrolled in a  Master's programme in Computer Science, Data Science, Artificial Intelligence, Machine Learning, or a related field
Solid grounding in machine learning fundamentals — supervised/unsupervised learning, model evaluation, feature engineering
Hands-on experience with ML frameworks such as PyTorch, TensorFlow, or scikit-learn (from coursework, research, or personal projects)
Programming proficiency in Python; experience with data manipulation libraries (pandas, NumPy, Spark is a plus)
Familiarity with NLP or embedding-based methods (e.g., Sentence Transformers, BERT-based models) is a strong plus
Interest in geospatial data, POI systems, addressing, or location intelligence
Analytical mindset with the ability to design experiments, interpret results critically, and communicate findings clearly
Collaborative and curious — comfortable asking questions, working iteratively, and learning from feedback

What you'll learn:

By the end of the internship you will have:
Worked on production-scale geospatial and POI data with real business impact
Gained experience in the full ML experimentation cycle - from problem framing and data analysis to model development and evaluation
Deepened your understanding of applied ML in a domain where data quality, scale, and semantic complexity are central challenges
Collaborated in a cross-functional team of scientists, engineers, and product managers