Machine Learning Research Intern (4–6 months)

Skypher, Inc.; Paris; Sur place; EUR 30 000 - 45 000; Plein temps

Avantages offerts par ce poste

Competitive compensation

Swile meal vouchers

Nice office in the 2nd Arrondissement

Partial remote work, up to two days per week

Technical mentorship

Résumé du poste

A leading AI company in Paris is looking for a Machine Learning Research Intern to contribute to applied ML projects. You will improve retrieval and reranking systems, build evaluation workflows, and develop pipelines for real-world data. This role values scientific rigor and an experimental mindset, with direct mentorship from experienced engineers. Ideal candidates should have Python skills and a background in quantitative fields. Partial remote work is offered, along with competitive compensation and meal vouchers.

Qualifications

• Currently pursuing or recently completed a degree in a quantitative field.
• Solid Python skills and a foundation in machine learning, deep learning, and NLP.
• Interested in cybersecurity or motivated to learn about security in tech companies.

Responsabilités

• Improving retrieval, reranking, and RAG pipelines for product use cases.
• Building evaluation workflows for model and system performance.
• Fine-tuning LLMs for focused tasks.
• Developing document understanding on real-world data.

Connaissances

Python skills

Machine learning foundation

Deep learning

NLP

English proficiency

Formation

Bachelor’s or Master’s degree in a quantitative field

Fourth year of higher education in engineering school

Outils

PyTorch

Git

AWS

Description du poste

Machine Learning Research Intern (4–6 months)

Paris / Remote Full Time

About Skypher

Skypher is a Y Combinator -backed AI agent platform that automates compliance questionnaires response, making it 10x faster. We save dozens of hours for many companies’ engineering, sales, customer success, and security leaders on security questionnaires and speed up their sales cycle by 20%.

Skypher was launched in 2020 in San Francisco and is now the market leader in security questionnaires response automation with 200+ global customers worldwide across North America, Europe and India. We work with Fortune 500 companies such as Edenred, Adobe, McKinsey & Company, TeamViewer or CMA CGM as well as fast growing tech companies like Deel, TCN.

Opportunity

We are looking for a Machine Learning Research Intern to join our growing engineering team and contribute to applied ML projects that are closely tied to the product.

You do not need prior cybersecurity or domain experience for this internship.

As an ML Research Intern, you will work on applied machine learning problems with direct product impact, alongside our Lead ML Engineer, CTO, and engineering team. Projects may include improving retrieval and reranking for RAG systems, building robust evaluation pipelines, adapting LLMs to targeted tasks, and finding innovative solutions to deal with complex, imperfect real‑world documents and data.

Beyond prototyping, you will be expected to approach problems with rigor: define meaningful baselines, run disciplined experiments, analyze failure modes, and learn how to turn the most promising ideas into clean, maintainable production code.

You will be working in an environment where thoughtful experimentation, creativity, and autonomy are genuinely valued.

Your responsibilities may include:

• Improving retrieval, reranking, and RAG pipelines for real product use cases
• Building evaluation workflows to assess model and system performance rigorously
• Fine‑tuning or adapting LLMs for focused, high‑value tasks
• Developing document understanding and data extraction pipelines on noisy real‑world data
• Studying model behavior through ablations, edge cases, and failure analysis
• Contributing production‑ready code and sound engineering practices

Profile

Must‑have

• You are currently pursuing, or have recently completed, a Bachelor’s or Master’s degree in a quantitative field such as Computer Science, Data Science, Statistics, Applied Mathematics, or Machine Learning
• Alternatively, you are studying in an engineering school and have reached at least the equivalent of a fourth year of higher education
• You have solid Python skills and a good foundation in machine learning, deep learning, and NLP
• You are interested in cybersecurity, or motivated to learn how security and compliance problems are addressed in modern technology companies
• You are comfortable working in English

Nice to have

• Experience with PyTorch, including training or implementing your own models
• Familiarity with Git and standard collaborative development workflows
• Some exposure to cloud platforms or CI/CD practices; AWS is a plus
• Familiarity with RAG systems, retrieval methods, or sentence similarity frameworks

What we value in candidates

We value candidates who enjoy understanding problems deeply, working carefully, and learning quickly in a product‑focused environment.

• Scientific rigor. You approach problems with structure, define clear baselines, choose meaningful metrics, and analyze results carefully.
• Experimental mindset. You aim to understand not only what works, but why — and why something may fail.
• Code quality. You write clean, well‑structured, maintainable code and care about producing tangible, usable outputs.
• Ownership and communication. You take responsibility for your work, communicate proactively, and build trust through reliability and follow‑through. You understand how your contribution fits into broader product and business goals.

What we offer

• Competitive compensation
• Swile meal vouchers
• Nice office in the 2nd Arrondissement
• Partial remote work, with up to two remote days per week
• Work on meaningful ML problems with close technical mentorship
• Learn how applied ML projects are scoped, evaluated, and shipped in a startup environment

Hiring process

• Initial call
• Technical data science take‑home exercise
• Technical interview with our machine learning team, including a debrief of the take‑home and a live coding exercise
• Final interview with one of the founders

To apply: send your application to [e-mail]