AutoML Research Internship

Owkin is hiring!



At OWKIN, we are dedicated to advancing the state-of-the-art in machine learning for medical research. From finding the next cancer-fighting molecule to tracking tumors, machine learning is an essential tool for pharmacologists, pathologists and many other medical practitioners and researchers. Our team is international, multidisciplinary with incredible talent in machine learning, medicine and business. Our data scientists are among the best in the world, with several Kaggle Masters, DREAM Challenge top performers, and publications in ICML, NIPS and other top scientific journals.
Owkin has raised a total of 18.1M$ with funds including Google Ventures and Otium capital, has offices in Paris, New-York, London and Nantes and is now looking for new talents.

OWKIN’s proprietary platform, OWKIN STUDIO, uses machine learning technology to integrate biomedical images, genomics and clinical data to discover biomarkers and mechanisms associated with diseases and treatment outcomes. OWKIN builds scientific collaborations with top-tier medical institutions and partners with leading pharmaceutical companies. OWKIN also develops state-of-the-art federated learning technologies in healthcare to overcome the need to share sensitive medical data, building collective intelligence from distributed data at scale, preserving privacy and security.

Job Description


We are looking for a passionate and motivated machine learning engineer intern to work with our research team.

The Machine Learning Engineer Intern is an intermediate between pure Data Science and pure Development.
Its role is to understand and implement Machine Learning algorithms and implement them at a production level.

This internship focuses on defining a robust AutoML algorithm for use on Owkin medical use-cases taking inspiration from the AutoML litterature (such as Architecture search, Bayesian Optimization, Neural Architecture search) and related fields (Meta-Learning, Domain adaptation, etc.)

This internship will involve:

  • Bibliography around AutoML and related fields
  • Definition of new AutoML algorithms,
  • Implementing AutoML algorithms, on a cluster.
  • Studies of the convergence time, guarantees
  • Consolidation of the algorithm metrics (convergence time, FLOPS*s), using theoretical guarantees, or heuristics

This work can lead to a publication, if successful.

AutoML is a rising field of research that aims to define algorithms and software architectures in order to find the best architecture to a given problem. For example, Neural Architecture Search algorithms have beaten the state of the art performance on the ImageNet challenge, in 2019 (top-5 accuracy).

AutoML is a promising technology in the industrial setting, because, it offers:
A gain of time for data scientists, avoiding the tedious hyperparameter search process
A gain of Cost for the company: Intelligent search can reduce drastically the number of computations, thus cost
Elaborate architecture space, that can raise models with performances way better than what can be found with manual search
New metrics to optimize architecture search, such as model size, training time etc.

However, AutoML, being an umbrella term grouping lots of different technologies (going from the basic grid search to the more advanced and costly Neural Architecture Search), needs to be specified for the specific problem to be solved.

More specifically, the Owkin medical data context poses an additional difficulty: Medical Imagery datasets propose relatively few samples (~500), but very high dimensional ones (~1Gb per sample): the resulting metrics have a huge variance, while training is still very long, having a real cost for the company.

The position is located in OWKIN Lab, our Paris-based R&D center. As such, the candidate must possess authorization to work legally in France. Candidates not meeting every criteria, but who can demonstrate exceptional skill in key areas, are invited to apply.

Preferred Experience


  • Currently enrolled in a MS or BS in applied mathematics, computer science, software engineering or an associated field
  • Python development expertise
  • Knowledge of production environments: Documentation / Specification, Unit tests, Deployment, Code optimisation
  • Knowledge and familiarity of machine learning fundamentals, from linear models to convolutional networks. A good grasp of concepts such as bias-variance tradeoff, under/overfitting, curse of dimensionality etc.
  • Experience with Machine Learning and Deep Learning frameworks (Tensorflow, PyTorch, etc.)
  • Development experience in Linux/Unix-like environments
  • Authorization to work legally in France


  • Participation in data science projects.
  • Contribution to publications at machine learning conferences and journals (NeurIPS, ICML, ICLR, etc.)
  • Knowledge of applied mathematics concepts (signal & image processing, statistics, etc…)
  • Experience with cloud computing services such as GCP, AWS, Azure, etc.
  • Knowledge with container-based deployments: Docker, Kubernetes, etc.
  • Experience developing solutions for distributed computation
  • Experience with system languages (C/C++) or scientific computing languages (Julia, R, etc.)
  • Experience with CUDA in C/C++
  • Contributions to OSS projects


  • ...enjoy working in a small, passionate team where your contributions will have a direct impact.
  • ...value process, standards, communication, and code quality.
  • ...want to help build never-before-seen technologies.


  • Competitive remuneration
  • International and High-profile team
  • The whole startup package: regular activities, nerf guns, onsite events...
  • Very convenient work atmosphere and access to premium computing hardware

Recruitment Process

The process is structured as follows:

  • Onsite technical interviews
  • Research presentation
  • Final interview

Owkin is an Equal Opportunity Employer. All qualified applicants will receive consideration for employment without regard to race, sex, gender, sexual orientation, age, color, religion, national origin, protected veteran status or on the basis of disability.

Additional Information

  • Contract Type: Internship (Between 6 and 6 months)
  • Location: Paris, France (75010)
  • Education Level: Master's Degree
  • Experience: < 6 months
  • Occasional remote authorized