LLM AI Engineer - Freelance (M/F)

France | Nov. 5, 2026

Report as Closed

Company: Collective.work

Country: France

Type: Onsite

Employment: Full-time

Description: Daily rate (TJM): 500 Within a team dedicated to technological platforms and data, the AI teams develop and operate large-scale artificial intelligence and machine learning solutions, for internal and external uses. The mission is part of a team in charge of developing a centralized and secure platform allowing the different teams to access artificial intelligence models: public LLM providers, sovereign European models and privately hosted models. This platform constitutes a point single entry point to integrate models and develop AI functionalities and agents. Its main challenges are: *    Unified access: single API, routing between models and providers, fallback mechanisms and load distribution. *    Governance and security: authentication, quotas, rate limiting, data protection and traceability. *    Cost control: consumption monitoring, budget management and usage analysis. *    Agentic ecosystem: integration of tool calling, MCP servers and flows linked to AI agents. The team adopts an AI-oriented development approach, with daily use of programming agents to accelerate developments. MISSIONS Development and operation of the AI ​​platform *    Design, develop and maintain the platform's functionalities: OpenAI compatible unified API, multi-provider routing, fallback, streaming and caching. *    Implement access control mechanisms, API key management, quotas, rate limiting and cost monitoring by team and use cases. *   Evaluate and compare models based on their quality, latency and cost. * Integrate new models and suppliers. * Develop observability: logs, traces, metrics, dashboards and alerting. *   Guarantee security and compliance: guardrails, personal data filtering, protection against prompt injections and traceability. *    Ensure the performance and reliability of the platform: SLO, load testing, capacity planning and incident management. Evolution towards agentic architectures *    Extend the platform to agentic uses: tool calling, exposure and governance of MCP servers. *    Expose AI services via APIs and MCPs. *    Provide SDKs, templates, plugins and documentation to facilitate adoption of the platform. Development augmented by AI *    Use AI development agents on a daily basis (Claude Code, Codex, Cursor, GitHub Copilot). *    Define specifications and divide developments into tasks that can be used by agents. *   Manage several agents in parallel and control the quality of their deliverables. *    Contribute to the improvement of development practices, automated testing and CI/CD pipelines. *    Participate in code reviews and technical documentation PROFILE WANTED *    Minimum 5 years of experience in software engineering. *   Strong skills in Python backend development (FastAPI, asynchronous programming) and API design. *    Expertise in AI/ML: machine learning, deep learning, operation of LLMs, embeddings, tokenization, fine-tuning and quantization. *    Hands-on experience in developing and operating LLM systems in production. *    Mastery of model provider APIs, streaming, tool calling and cost and token management. *   Hands-on experience developing with AI agents. *    Good understanding of agentic protocols and frameworks: MCP, A2A, LangGraph, Google ADK or equivalent. *    Cloud Native experience: Docker, Kubernetes, Terraform, CI/CD and Google Cloud Platform. *    Security knowledge: OAuth2/OIDC, authentication, authorization and

Apply here:

Web: Apply here

Emails: