Senior AI Platform Engineer (on-site)

Job description

At Sword Technologies, we support our clients in their digital transformation projects by contributing to high value-added missions within demanding environments. We place particular emphasis on the quality of the assignments we offer, as well as on supporting our employees in their professional development. As part of our growth, we are looking for talented individuals to strengthen our teams. We are looking for a Senior AI Platform Engineer able to maintain and evolve an enterprise AI platform in a regulated environment, combining production software engineering, technical governance and security by design. Mission: As part of an assignment with a large European institutional client in the financial sector, you will act as a senior external consultant to ensure the operation, maintenance, security and controlled evolution of an enterprise AI platform in production. This platform provides governed access to large language models (LLMs), knowledge bases, an enterprise search engine, conversational agents, evaluation services and an administration console. You will work in close collaboration with the client's application, infrastructure, cybersecurity, data and business teams. On a daily basis, you will maintain and enrich the Python/FastAPI gateway, the agent engine, policy controls, connectors, background processing, the PostgreSQL layer and the administration console. You will operate the Retrieval-Augmented Generation (RAG) flows and knowledge integrations — including enterprise search, SharePoint and the intranet. You will design and maintain reproducible evaluation workflows for AI agents, covering quality, compliance, costs and regression detection. You will implement and oversee safeguards against prompt injection attacks, confidential data leaks, personal data (PII) handling and adherence to internal policies. You will handle production incident management — latency diagnostics, scalability, token consumption — and maintain secure authentication and authorization based on OIDC/OAuth2, role-based access control (RBAC) and multi-tenant isolation. You will oversee containerized deployments, CI/CD pipelines, infrastructure-as-code (Terraform, Helm, Kubernetes) and production observability via OpenTelemetry, Splunk and Prometheus. You will produce and maintain automated tests, runbooks, architecture decisions and operational documentation, and ensure a structured knowledge transfer to the client's internal teams. Technical environment: Must have: Solid experience in production software engineering with modern Python, asynchronous programming, FastAPI, Pydantic and typed application design In-depth mastery of PostgreSQL: SQL, transactional processing, schema design, forward-only migrations, concurrency and data integrity Hands-on experience with LLMs, agent orchestration, RAG, embeddings, enterprise search, prompt design, tool calling and model abstraction Experience designing AI evaluation frameworks covering quality, groundedness, security, regression detection, latency, token consumption and costs Understanding of AI security threats and controls: prompt injection, attacks on retrieved content, data leakage, insecure tool execution and handling of sensitive data Mastery of API security: OIDC/OAuth2, JWT validation, role- and policy-based access control, secrets management and secure development practices Experience with Docker, Kubernetes, Helm, Terraform, CI/CD and operating applications in Azure or a comparable enterprise cloud environment Experience with production observability: structured logging, metrics, distributed tracing, alerting and tools such as OpenTelemetry, Splunk and Prometheus Strong skills in automated testing: unit, integration, database, API, security and frontend Working knowledge of TypeScript and React sufficient to maintain and evolve a web administration console Experience with Git-based delivery: code review, static analysis, dependency management and quality gates Minimum 5 years of professional experience in software engineering or platform engineering with responsibility for production systems subject to security, availability or regulatory compliance requirements Nice to have: Experience with Azure AI services, Azure OpenAI, Azure AI Search, Microsoft Graph or SharePoint integrations Experience with the governance of the AI model and agent lifecycle in a production environment Knowledge of the EU AI Act, GDPR, DORA, financial-sector ICT controls or comparable governance frameworks Experience in load testing, performance engineering, capacity planning, disaster recovery and business continuity exercises Familiarity with the Model Context Protocol (MCP) or comparable tool integration standards Experience supporting multilingual enterprise applications Experience with Temporal for workflow orchestration Proficiency in French; professional English is required Prior experience