Mlops Scientist

Fraunhofer Portugal Aicos; Porto; Hybrid; EUR 55,000 - 75,000; Full time

Benefits offered by this job

Partial health insurance

Flexible working hours

Hybrid work

Job summary

Fraunhofer Portugal AICOS is seeking an MLOps Scientist to join the Intelligent Systems Group, focusing on machine learning and health data analysis. You will develop privacy-preserving ML pipelines and contribute to a flexible distributed software framework supporting federated learning and data governance.

Your role includes following MLOps best practices, operating federated systems with IAM and monitoring, and implementing data governance across datasets and models; collaboration with

Qualifications

• Master's degree or equivalent in AI, AI & Data Science, Biomedical Engineering, CS & Engineering or similar.
• Minimum 18 months of relevant work experience in research departments in industry, applied research institutions, or universities.

Responsibilities

• Follow MLOps infrastructure and deployment best practices for robust ML applications, including containerization and production monitoring.
• Operationalize federated learning systems with identity and access management, client orchestration, training coordination and lifecycle management.
• Establish rules for data access, sensitivity classification, retention schedules, and audit trails.
• Manage a data governance tool for monitoring datasets and AI models under development and testing.

Skills

Python

ML frameworks

English proficiency

APIs design

DVC & MLflow

Federated Learning

Education

Master's degree in AI/CS/ biomedical engineering

Tools

Cloud platforms (AWS/GCP/Azure)

Data governance tools

Job description

Fraunhofer Portugal AICOS is seeking a MLOps Scientist to join our dynamic and multidisciplinary team within the Intelligent Systems Group, namely in machine learning and tabular data analysis applied to health data. It is expected that you will contribute to the business areas of Digital Health and Digital Futures.

Specifically, in this job will participate in the research and development of methodologies to improve the privacy and robustness of Machine Learning (ML) models. You will contribute to improving a flexible distributed software framework, which includes federated learning and data governance to support the deployment of privacy-preserving ML pipelines.

Your role:

Follow MLOps infrastructure and deployment best practices for robust ML applications, including containerization, and production monitoring to ensure federated learning models run reliably at scale;

Operationalize federated learning systems with identity and access management, client orchestration, training coordination, monitoring, and lifecycle management across partners;

Establish clear rules for data access, classification by sensitivity, retention schedules, and audit trails;

Manage a data governance tool for monitoring active datasets and AI models under development and testing, to support AI workflows with predictive modelling and classification of private healthcare data;

Supervise of datagovernance strategies for ML with federated data catalogues, including dataset cards, federated model card, ML experiment tracking, and data lineage;

Implement automated quality checks, validation rules, and ongoing monitoring to keep data accurate and qualified;

Work closely withmultidisciplinary teams to:

Maintain and refine current codebases, adhering to best practices in software development and reproducible research;

Promote and integrate data governance best practices into existing software pipelines or platforms;

Contribute to strategic discussions on advancing research directions, staying updated on cutting-edge MLOps frameworks and methodologies related to privacy-preserving and trustworthy AI.

Your profile: Mandatory Requirements:

Academic Qualifications

: Master's degree or equivalent academic qualifications in Artificial Intelligence, Artificial Intelligence and Data Science, Biomedical Engineering, Computer Science and Engineering or similar.

Professional Experience

: Minimum of 18 months of relevant work experience in research departments in industry, applied research institutions, or universities.

Technical Skills:

Experience in API design and implementation, as well as deployment of ML systems in diverse environments (on-prem, cloud, edge);

Programming proficiency in Python and experience with machine learning frameworks such as Scikit-learn, PyTorch or TensorFlow;

Experience with Data versioning (DVC) and ML experiment tracking (MLFlow);

Experience in privacy-preserving ML techniques such as Differential Privacy, Homomorphic Encryption, Federated Learning or Secure Multi-Party Computation.

Other Skills:

Proficient in English, preferably with a C2-level in the CEFR (Common European Framework of Reference for Languages);

Excellent communication skills for technical and general audiences;

Capacity to translate research findings into actionable insights for real-world applications.

We value:

Familiarity with MLOps practices and frameworks, including data governance and ML model registry tools;

Experience with cloud platforms such as AWS, GCP, or Azure;

Experience in development and implementation of RAG Systems;

Experience in implementation and comparison of Federated Learning strategies;

Experience in rawcoding and vibe coding, balancing a curiosity-driven approach with software engineering principles;

Strong interest in security and privacy considerations within distributed ML systems;

Commitment to clear and honest communication of data limitations using established reporting standards;

Autonomous, dependable, proactive, and a critical-thinking team player.

Why should you join Fraunhofer Portugal:

Innovative Environment: Be part of a people-centric workplace that fosters creativity and out-of-the-box thinking. We encourage the development of new ideas and ensure that every voice is heard;

Research with Impact: Engage in projects that sit at the intersection of research and real-world applications, contributing to technology that makes a tangible difference in society;

MultidisciplinaryTeams: Collaborate with professionals from diverse backgrounds, enhancing your learning and professional growth;

Professional Excellence: Work within a culture that upholds professional standards and best practices, promoting continuous improvement and excellence in research;

Flexible Work Arrangements: Benefit from flexible working hours and hybrid work opportunities, supporting a healthy work-life balance;

Comprehensive Benefits: Benefit from a partially funded health insurance plan, and a variety of additional perks;

Supportive Culture: Join a team with an excellent spirit, where collaboration, mutual support, and team achievements are celebrated.

Selection process: Admitted candidates (those who fulfill the mandatory requirements and application process) will be subject to a curricular evaluation based on the predefined evaluation criteria and weights.

The evaluation is based on a quantitative scoring system ranging from 0 to 100 points, structured as follows:

Curricular Evaluation – 70%;

Interview – 30%;

The curricular evaluation will assess the mandatory and preferential requirements, as follows:

Professional Experience: 20%;

Experience in APIdesign and implementation, as well as deployment of ML systems in diverse environments (on-prem, cloud, edge): 20%;

Programming proficiency in Python and experience with machine learning frameworks: 10%;

Experience with Data versioning and ML experiment tracking: 10%;

Experience in privacy-preserving ML techniques such as Differential Privacy, Homomorphic Encryption, Federated Learning or Secure Multi-Party Computation: 10%;

Other Skills: 10%.

Based on the results of the curricular evaluation, the top-ranked candidates (typically up to three) will be invited for an interview. Only candidates achieving a minimum score of 70 points in the curricular evaluation will be considered for progression.

The interview will be conducted in accordance with the predefined evaluation criteria and will contribute to the final classification. The interview will assess candidates across the following dimensions:

Motivation and alignment with the role and organizational values: 20%;

Technical knowledge and professional experience relevant to the position: 20%;

Problem-solving capacity: 20%;

Analytical skillscapacity: 20%;

The interview scoring will be conducted using a structured evaluation matrix to ensure consistency and comparability between candidates.

For positions requiring an additional interview stage, only candidates achieving a minimum score of 80 points in the first interview will proceed to the next stage.

The final classification of candidates will be based on the results of all evaluation stages, in accordance with the defined weights.

All applicants will be notified of the results of the call via the email address provided in their application.

After notification, all candidates have 10 days to comment.

Selection Panel: -Filipe Soares (PhD;
Permanent Member).

Non-discrimination and equal opportunity policy: Associação Fraunhofer Portugal Research actively promotes a non-discrimination and equal opportunities policy, ensuring that no candidate can be privileged, benefited, impaired or deprived of any rights whatsoever, or be exempt of any duties based on their ancestry, age, sex, sexual orientation, marital status, family and economic conditions, instruction, origin or social conditions, genetic heritage, reduced work capacity, disability, chronic illness, nationality, ethnic origin or race, origin territory, language, religion, political or ideological convictions and union membership.

To comply with Law no. 4/2019, of 10 January, candidates must declare on the application form, under a statement of honor, their degree of disability, the type of disability and the means of communication/expression to be used in the selection process.

Privacy Policy: All personal data collected during the recruitment process will be handled in accordance with applicable data protection regulations.

For more information on how personal data is processed, please refer to the Associação Fraunhofer Portugal Research Privacy Policy available on the institutional website Privacy Policy.

Note: The research activities are planned to be developed within the following projects:

AIFH – Artificial Intelligence Factory in Health, Reference No. 26549;

FEDAS – FederatedData Spaces for Privacy-Preserving Data Valorisation, and Project Reference No. 28795.