AI Engineer

Publicis Groupe Holdings B.V; Paris; Sur place; EUR 70 000 - 110 000; Plein temps

Avantages offerts par ce poste

Télétravail partiel

Travail international possible (Work‑Y

Formation continue & certifications

Résumé du poste

Publicis Groupe recherche un ingénieur IA pour rejoindre l’équipe Publicis Production AI et construire la prochaine génération de produits et workflows basés sur l’IA générative pour la création de contenu et de campagnes.

Vous combinerez ingénierie back-end, AI appliqué et déploiement dans des environnements cloud (Azure/GCP), en favorisant la qualité production, l’évolutivité et les interfaces avec les équipes produit et design.

Qualifications

• Master ou équivalent en informatique, IA ou domaine lié.

Responsabilités

• Concevoir et maintenir des API backend évolutives avec FastAPI.

Connaissances

Python

FastAPI

API design

Git & CI/CD

Cloud platforms

LangChain/LangGraph

Evaluation & prompt engineering

Formation

Master en informatique/IA

Outils

OpenAI

Gemini/Vertex AI

Docker

Azure

GCP

Description du poste

Description de l'entreprise

Within Publicis Groupe’s Intelligent Creativity business, we specialize in bringing creative ideas to life, and to consumers.

By combining 100 years of craft excellence with 6,000 experts across 52 locations of the world’s biggest studio network, we leverage the industry’s richest data, through the power of agentic AI, to radically redefine content production with Intelligent Content. We intuitively deliver this through Marcel Make, the world’s first Intelligent Content agent. The result? Predictively performing content that unlocks business growth in unprecedented ways.

No more guesswork. No more waste. Just content that works, working a lot harder.

The Publicis Production AI team leads innovation in generative AI, developing and deploying large language models (LLMs) and advanced AI technologies for business impact.

Descriptif du poste

We are looking for an AI Engineer to join the Publicis Production AI team and help build the next generation of AI-powered products and workflows for content and campaign creation. The engineer will join a multidisciplinary product team and work closely with AI engineers, software engineers, product managers, designers, and production specialists to turn emerging Generative AI capabilities into reliable, scalable products used by internal teams and clients.

The role is strongly hands‑on and combines backend engineering with applied AI. You will design and build scalable Python services and APIs using FastAPI, orchestrate large language, image, video, audio, and multimodal models, and develop innovative agentic and GenAI pipelines that automate or augment key stages of the campaign creation process. This includes workflows such as brief understanding, ideation, content generation, adaptation, quality checks, brand compliance, localization, and production automation.

You will work in a fast-moving environment where experimentation is important, but production quality matters equally. The objective is not only to prototype new AI capabilities, but to engineer them into secure, observable, maintainable services that can operate at scale across brands, markets, users, and production workflows.

Responsabilités

• Design, develop, and maintain scalable backend APIs and microservices in Python, primarily using FastAPI, to expose AI capabilities to Publicis Production products and internal platforms.
• Build orchestration layers that connect and manage multiple Generative AI models and providers, including LLMs, image generation, video generation, audio, embeddings, and multimodal models.
• Design and implement agentic workflows using modern orchestration patterns, including tool calling, structured outputs, state management, routing, retries, memory, and human-in-the-loop validation.
• Create innovative end-to-end GenAI pipelines for advertising and campaign creation, from brief interpretation and creative ideation through generation, adaptation, localization, validation, and production-ready outputs.
• Develop reusable AI components and services that can be integrated across multiple products, clients, markets, and use cases rather than building one-off prototypes.
• Integrate third‑party AI APIs and SDKs such as OpenAI, Google Gemini/Vertex AI, Adobe Firefly, Anthropic, and other emerging model providers, while abstracting provider‑specific implementation details where appropriate.
• Implement Retrieval‑Augmented Generation (RAG), embeddings, vector search, structured data retrieval, and grounding approaches when applications require access to brand, campaign, product, or operational knowledge.
• Evaluate models and prompting strategies for quality, latency, reliability, safety, and cost, and translate experimental findings into engineering decisions for production systems.
• Build robust data‑processing and asynchronous workflows for handling documents, images, video, and other creative assets at production scale.
• Implement production engineering best practices including automated testing, API validation, error handling, logging, monitoring, tracing, rate limiting, caching, and performance optimization.
• Deploy and operate AI services in cloud environments, primarily Azure and Google Cloud Platform (GCP), and contribute to CI/CD, containerization, configuration, secrets management, and production support.
• Collaborate with frontend, platform, DevOps, product, and data teams to define technical solutions and integrate AI services into user‑facing products.
• Participate actively in agile product delivery through technical design, sprint planning, code reviews, pull requests, documentation, troubleshooting, and continuous improvement.
• Stay current with rapidly evolving Generative AI technologies, test new models and frameworks, and identify opportunities to turn promising research or vendor capabilities into practical production features.

Qualifications

• Master’s degree in Computer Science, Engineering, Mathematics, Artificial Intelligence, or a related technical field, or equivalent practical experience.
• Strong Python engineering skills and experience building production‑quality backend applications or services.
• Hands‑on experience designing REST APIs, ideally with FastAPI, including request validation, asynchronous processing, authentication patterns, error handling, and API documentation.
• Good understanding of software engineering principles including modular architecture, object‑oriented and functional programming patterns, testing, maintainability, and code quality.
• Practical understanding of Generative AI concepts including LLMs, agents, RAG, embeddings, structured outputs, tool/function calling, multimodal models, text‑to‑image, and diffusion‑based generation.
• Experience using APIs or SDKs from leading AI providers such as OpenAI, Gemini/Vertex AI, Claude, Adobe Firefly, or equivalent platforms.
• Experience with agent or workflow orchestration frameworks such as LangGraph, LangChain, Semantic Kernel, or equivalent patterns is a strong advantage.
• Familiarity with cloud platforms, preferably Azure or Google Cloud Platform (GCP), and common cloud‑native concepts such as containers, managed services, queues, databases, object storage, and secrets management.
• Strong knowledge of Git and collaborative development practices including branching, pull requests, code reviews, and CI/CD workflows.
• Ability to move comfortably between experimentation and production engineering, with a strong interest in measuring quality, performance latency, reliability, and cost.
• Strong problem‑solving skills, curiosity, and the ability to learn quickly in a rapidly evolving technical domain.
• Ability to collaborate effectively with engineers as well as product, design, creative, and production stakeholders in an international team.

Nice to Have

• Experience building AI‑powered products, workflow automation, creative technology, or content‑generation systems.
• Experience with vector databases, PostgreSQL, Redis, message queues, background workers, or event‑driven architectures.
• Experience with Docker, Cloud Run, or similar containerized deployment environments.
• Experience processing creative assets such as images, video, audio, PDFs, or design files.
• Knowledge of AI evaluation, guardrails, content safety, prompt/version management, observability, and responsible AI practices.
• Interest in advertising, creative production, personalization, localization, or marketing technology.

Informations complémentaires

What You Will Work On

You will contribute to AI products and capabilities used to support real advertising production workflows. Depending on the product team, this can include intelligent brief processing, creative ideation, agent‑driven workflow automation, image and video generation, scalable content adaptation, localization, brand and compliance checks, asset analysis, and integrations with production platforms and external AI providers.

Rejoindre Publicis Groupe, c’est :

• Pouvoir télétravailler jusqu'à 2 jours par semaine
• La possibilité de travailler dans le monde entier jusqu’à 6 semaines continues chaque année à partir d’un an d’ancienneté (Programme Talents Work Your World)
• Se former tout au long de son parcours professionnel grâce aux 30.000 modules accessibles en e‑learning et comprenant des certifications en partenariat avec LinkedIn Learning sur la plateforme propriétaire Marcel

Viva La Différence !

Cette philosophie de Publicis Groupe témoigne depuis toujours de notre engagement pour la diversité et de la conviction que nos talents sont notre plus grande richesse et notre meilleur atout.

Nous valorisons ainsi toutes les singularités, sans distinction d’âge, de sexe, de couleur de peau, d’origine sociale, de religion, ou d’orientation sexuelle... seules la compétence et l’énergie comptent ! Nous encourageons toutes les candidatures qualifiées et seront ravis d’accompagner tout au long du processus de recrutement, de manière personnalisée un.e candidat.e en situation de handicap qui en ferait la demande.

Publicis France est engagé pour l’égalité des chances et l’équité d’opportunités pour tous et toutes.