Ingénieur en science des données Intelligence Artificielle

About IBM Consulting

A career at IBM Consulting is built on lasting client relationships and close global collaboration. You'll work with leading companies across all industries, helping them define their hybrid cloud and artificial intelligence strategy. With the support of our strategic partners, robust IBM technology, and Red Hat, you'll have the tools you need to drive meaningful change and accelerate impact for our clients. At IBM Consulting, curiosity is the engine of success. You'll be encouraged to challenge the status quo, explore new ideas, and create innovative solutions that deliver tangible results. Our culture of growth and empathy focuses on your long-term career development while valuing your unique skills and experiences.

Your Role And Responsibilities

In a context of accelerating AI usage, we are looking for an experienced AI Data Science Engineer capable of intervening across the entire AI product lifecycle, from identifying business opportunities to industrialization and production monitoring. This role covers classic machine learning as well as deep learning, generative AI, agentic AI, multimodal models, model evaluation, data engineering for AI, and the production of robust, scalable, and measurable solutions.

The position is highly cross-functional and sits at the intersection of R&D, delivery, data, product, and business. The expected candidate must be able to quickly prototype, structure rigorous experimental approaches, industrialize reliable pipelines, and guide the team's technological choices through active monitoring and strong analytical skills.

• Identify, frame, and prioritize high-impact AI use cases, in line with business objectives, technical constraints, and operational feasibility.
• Design and develop machine learning, deep learning, and generative AI solutions on structured, textual, visual, documentary, or multimodal data.
• Prepare, enrich, and qualify data necessary for models: collection, cleaning, labeling, augmentation, synthesis, anonymization, quality control, and versioning.
• Design robust experimentation protocols, define relevant metrics, conduct benchmarks, analyze results, and formulate reasoned recommendations.
• Fine-tune, adapt, or orchestrate existing models, including LLMs, VLMs, and specialized models, according to use cases and constraints of cost, latency, robustness, and explainability.
• Develop and industrialize training, evaluation, inference, and monitoring pipelines in collaboration with data and engineering teams.
• Deploy AI models and services in a production environment, ensure their observability, monitor their drift, measure their business performance, and drive their continuous improvement.
• Implement and disseminate standards for reproducibility, quality, traceability, documentation, and governance of experiments.
• Conduct active technological monitoring of emerging models, frameworks, tools, benchmarks, and practices, then transform this monitoring into concrete decisions, POCs, or architectural recommendations.
• Support junior team members, challenge technical choices, contribute to good development practices, and actively participate in structuring decisions on AI stacks.

Required Technical And Professional Expertise

• Excellent command of Python and major data science and deep learning frameworks, including scikit-learn, PyTorch, Hugging Face, pandas, NumPy, OpenCV, and TensorFlow.
• Solid experience in supervised and unsupervised machine learning on tabular, behavioral, or business data: classification, regression, clustering, scoring, ranking, anomaly detection, feature engineering, and interpretability.
• Hands-on experience with LLMs/VLMs: prompting, evaluation, RAG, light or full fine-tuning, model adaptation, guardrails, quality measurement, and cost optimization.
• Mastery of agentic AI frameworks: LangGraph, LangChain, LlamaIndex, AutoGen, Semantic Kernel, or equivalent frameworks, with the ability to design robust, traceable, and maintainable workflows.
• Ability to design reliable agents in an enterprise context: permission management, securing tool calls, output validation, human-in-the-loop, failure management, monitoring, auditability, and cost control.
• Mastery of Document AI challenges: OCR, parsing, information extraction, document classification, layout analysis, table extraction, structuring complex content.
• Very good command of deep learning applied to NLP, computer vision, documents, and multimodal approaches.
• Good command of experimentation and reproducibility pipelines: MLflow, Weights & Biases, DVC, Git, industrialized notebooks, testing, packaging, and CI/CD for AI.
• Good understanding of AI infrastructure: training and inference on GPUs, containerization, serving APIs, inference optimization, quantization, monitoring, and resource management.
• Ability to work with cloud or on-prem environments, data pipelines, and production architectures integrated into the IS.
• Excellent communication in French and English, both written and spoken.

Preferred Technical And Professional Experience

• Ability to transform a vague need into a framed data science problem, with hypotheses, success criteria, experimentation plan, and production roadmap.
• Ability to arbitrate between model performance, operating cost, maintainability, time-to-market, technical debt, and operational risks.
• Excellent technical communication, both to expert profiles and to product, business, or management stakeholders.
• Technical leadership, strong autonomy, ability to mentor or upskill other data/AI profiles.
• Critical thinking about benchmarks, data quality, biases, robustness, security, ethics, and the real value of proposed solutions.