Data Scientist IA Generativa

We are looking for the person who can turn playing with LLMs into designing generative AI products that work, scale, and improve how work is done in the real world. Who are we? At PROCESIA we build solutions where generative AI stops being hype and becomes assistants, intelligent search engines, automations and services that are used every day. The challenge: Data Scientist Generative AI This role is about designing how AI thinks within our products: what information it consults, how it combines it, how it responds, and how we improve its behavior with real data. It's not just a job of training models; it's an impactful role, very closely tied to the product, architecture, and user experience. What we would like to find in you: A solid track record in IT projects (around 7 years or more) and several of them focused on data and artificial intelligence. At least 5 years participating in analysis, design and development of systems that work with large volumes of information or AI solutions. Complete confidence in Python and its data ecosystem: pandas, numpy, scikit-learn, visualization with matplotlib/seaborn and intensive use of notebooks. Real-world experience building APIs and services around models (FastAPI/Flask, OpenAPI/Swagger, CI/CD, Docker, Kubernetes). Several projects under your belt with generative AI : RAG architectures, GPT-type models, LLaMa or Mistral, Transformers, fine-tuning, embeddings and metrics to evaluate LLMs. What will your day-to-day life be like? Instead of being in a data tower, you'll work very closely with development and architecture, from idea to deployment. You will design RAG-based generative AI solutions: deciding how the index is built, what context is retrieved, and how the LLM call is orchestrated. You will prepare and understand the data (structured and unstructured), defining the transformations they need to feed the generative models. You will test, compare and adjust large models (GPT, LLaMa, Mistral), defining prompt strategies, fine-tuning and continuous evaluation. You will work hand in hand with the Python team to expose your models through robust, monitored APIs deployed in cloud-native environments. You will measure the impact of what you do: quality of responses, times, user feedback, and opportunities for improvement. Beyond meeting minimum requirements, we will focus on: Real-world cases where you have used generative AI to solve specific problems (not just PoCs that remained in a repo). How do you combine your technical profile with the ability to explain decisions and results to non-technical people? Additional experience moving comfortably between containers, Kubernetes, observability (Grafana, Prometheus) and vector databases. What you will find at PROCESIA We want this role to be a qualitative leap in your career, not just another project. A stable project where generative AI is the core, not an ornament, and with a medium-to-long-term trajectory. Permanent contract, full-time, hybrid/remote model and flexible hours so you can combine deep focus with personal life. Career plan designed with you, English classes and real support for technical certifications that interest you. 3 months of intensive work schedule in summer, intensive every Friday and Christmas Eve and New Year's Eve off. Health insurance without co-payments, a friendly atmosphere, zero egos, free coffee and tea in the office, and a Management 3.0 model where decisions are made collaboratively. Competitive salary aligned with the value you bring to the project.