Senior Deep Learning Scientist

About the role

You will own the design, development, and productionization of deep learning and computer vision algorithms that operate on complex real-world data such as high-resolution satellite images, FMV video streams, and acoustic signals. You will be responsible for translating ambiguous operational requirements into robust technical solutions that deliver measurable value in defense, intelligence, and aerospace contexts. Within the AI Production department, you will act as a key technical contributor and thought partner, shaping the architecture and quality of algorithms that directly support mission-critical applications. You will collaborate daily with multidisciplinary teams, ensuring that your technical work aligns with strict operational, security, and performance standards. This role requires a high level of ownership, from initial feasibility studies through to scalable deployment and continuous improvement in production environments.

Key facts

What you'll do

Conduct in-depth analysis of customer problems and translate them into technical specifications for deep learning and computer vision solutions.

Design, develop, train, and evaluate algorithms for object detection, classification, and tracking across satellite imagery, video, and acoustic data streams.

Implement and refine end-to-end machine learning pipelines that integrate data curation, feature engineering, model training, and performance validation.

Champion the adoption of the group's internal AI framework to ensure consistent, reproducible, and high-performance model development across teams.

Drive the industrialization of algorithms by leveraging and improving MLOps tools, focusing on scalability, reliability, and monitoring in production.

Perform regular technical reviews and contribute to the definition of software architecture, data schemas, and integration interfaces for AI components.

Establish and promote best practices for data management, experiment tracking, versioning, and model governance to improve team efficiency and auditability.

Support commercial and technical sales activities by providing technical explanations, scoping workshops, and proof-of-concept demonstrations during pre-sales phases.

Mentor junior data scientists and engineers by reviewing their work, facilitating knowledge transfer, and elevating the overall technical bar of the team.

Stay up to date with advances in deep learning research and adjacent fields, and assess the applicability of new techniques to Safran.AI's operational problems.

Contribute to the definition and execution of the technical roadmap for the AI Production department, including toolchain enhancements and platform evolution.

Collaborate with AI Platform and other departments to align on standards, reusable components, and cross-cutting concerns such as security and compliance.

Ensure that developed solutions meet stringent operational constraints related to performance, robustness, and safety for defense and aerospace applications.

Act as a technical interface between data science, software engineering, and domain experts to ensure smooth delivery of integrated solutions.

Requirements

You hold a degree in Software Engineering, Deep Learning, Computer Vision, or a closely related technical field.

You bring more than 6 years of professional experience in deep learning and/or computer vision, ideally applied to domains such as satellite imagery, video, industrial 4.0, or acoustic signal processing.

You have hands-on experience with production environments, including deployment, monitoring, and operational observability of machine learning systems.

You possess strong proficiency in Python and adhere to rigorous software development best practices in your daily work.

You have solid software engineering skills, including system architecture, design patterns, testing, and CI/CD workflows.

You demonstrate a solid understanding of data storage, orchestration frameworks, and containerization technologies.

You are experienced working in agile environments while maintaining flexibility and avoiding dogmatic adherence to any single methodology.

You can communicate complex technical ideas clearly and defend architectural choices to both technical and non-technical stakeholders.

You exhibit a user-centric mindset, prioritizing user needs and operational constraints in your design decisions.

You show openness to collaboration, active listening, and effective teamwork across cultural and functional boundaries.

You are curious and motivated to learn technically, whether through data science methods, operational processes, or complementary disciplines.

You are willing to obtain the mandatory defense clearance required for this position.

Nice to have

Experience with satellite, aerial, or geospatial data applications.

Familiarity with acoustic or signal processing workflows relevant to defense or surveillance.

Knowledge of MLOps platforms, experiment tracking, and model monitoring tools.

Understanding of safety-critical or regulated environments and their impact on AI system development.

Practical notes

This position is based in Paris and is open to full-time employees.

Candidates must be able to obtain a defense clearance to access classified projects and environments.

The role may involve occasional travel within France and Europe for project meetings or industrial visits.

Visa sponsorship may be considered for eligible international candidates depending on legal and administrative requirements.

Application deadlines are aligned with rolling recruitment processes, and early submission is recommended due to the high volume of qualified candidates.