Data Scientist

Overview

Data Scientist - 12 Months Contract - Central Paris, France

Hybrid 3 days on site, 2 days remote

Must be able to speak fluent French and a good command of English

CVs need to be in English

October start date

Essential qualifications

• Successful experience of at least 5 years in Data Science. Expert in Python. Generative AI. Expert in fundamental concepts of deep learning and Transformer architecture.
• Expert demonstrated ability to conduct experiments and rigorously measure results.
• Expert strong ability to analyze and rapid prototyping on LLMs-related issues (RAG, Fine-tuning, Prompt Engineering, Multi-Agents, etc.).
• Expert Git and DevOps context (CI/CD, with GitHub or similar).
• Expert mastery of key frameworks in the ecosystem: LangChain, LlamaIndex, Transformers, CrewAI, etc.
• Appetite for craftsmanship / clean code software; environment (squad, CoP, coaches) helps to propagate.

Objectives and deliverables

My client is looking for an LLM Data Scientist Engineer with experience in Generative AI.

Data Science consists of extracting, analysing and valorising data from various sources, in order to solve business problems or create new opportunities. The skills required to provide a Data Science service are as follows:

• Prototype and evaluate agentic AI approaches, where the LLM uses tools (APIs, databases, code) to plan and execute complex multi-step tasks.
• Conduct targeted fine-tuning experiments (e.g. LoRA) to specialize them on specific data and business tasks and measure their qualitative contribution.
• Perform a rigorous comparative analysis of the different models and techniques (RAG vs. fine-tuning vs. agents) and frameworks on the market to guide technology choices.
• Design and test advanced workflows combining several techniques (e.g. RAG for knowledge, fine-tuning for style, agents for action) to solve complex problems.
• Explore the reasoning capabilities of models (Chain-of-Thought, ReAct, etc.) and evaluate their reliability for analytical or decision-making use cases.
• Produce clear deliverables: functional demonstrators, benchmarking reports, and recommendations on the viability and applicability of the solutions explored.

Cross-cutting contributions

• Participate in the design/maintenance of new tools/packages common to the organization.
• Challenge the team's vision.
• Participate in CoPs (Communities of Practice) and team rituals.

Skills required

To set up a development in data science and AI, you need to have the following skills:

• Must have: Proficiency in the fundamental concepts of Deep Learning and Transformer architecture. Strong ability to analyse and rapid prototyping on issues related to LLMs (RAG, Fine-tuning, Prompt Engineering, Multi-Agents, etc.). Mastery of key frameworks in the ecosystem: LangChain, LlamaIndex, Transformers, CrewAI, etc. Strong command of the Python language and its environment. Demonstrated ability to conduct experiments and rigorously measure results. Soft skills: curiosity, creativity, rigor, critical thinking, business sense.
• Software development skills required: Must have: a good command of git and a first experience in a DevOps context (CI/CD, with GitHub or similar), an appetite for craftsmanship / clean code software, which the TDF environment (squad, CoP, coaches) helps to propagate. Soft skills: collaboration, communication, autonomy, adaptability, problem solving.