Site Reliability Engineer – AI Applications (M/W/X)

Description

The company is rethinking how its teams work with information using AI, building and running applications for document-heavy, interactive and intelligent use cases such as hybrid search, question answering, document generation, conversational assistants, agentic workflows and coding assistants.

The role is on the team that keeps the AI platform reliable, scalable and well run. That platform includes the AI Gateway, MCP servers, LLM serving infrastructure, retrieval services and the supporting platform components that let teams take AI applications to production safely.

You will work where fast AI engineering meets disciplined site reliability engineering, helping the platform change quickly while keeping it observable, resilient, cost-efficient and dependable at scale.

Responsibilities

• Run and improve the reliability, scalability and performance of the AI platform, covering the AI Gateway, MCP servers, LLM serving, retrieval services and related components.
• Set and maintain SLOs, SLIs and error budgets, and use them to balance release speed against stability.
• Build observability across infrastructure, services and AI-specific signals, including latency, token throughput, model and provider errors, cost, saturation and quality.
• Make model and API serving faster, more efficient and cheaper across cloud, containerized and high-throughput environments.
• Design graceful degradation, fallback and failover, including resilience at the model, provider and region level.
• Keep deployments safe and repeatable using GitLab CI/CD, infrastructure as code, automated testing and progressive delivery where it fits.
• Lead or take part in incident response, post-incident reviews, capacity planning and reliability work.
• Work closely with software engineers, data scientists, ML engineers, security teams and product stakeholders to put AI services into production responsibly.
• Help set operational standards, runbooks, dashboards and practices for AI application reliability.

Requirements

• You have proven experience in SRE, platform, backend or infrastructure roles in compute-intensive, data-intensive or distributed environments.
• You understand reliability engineering well, including SLOs, error budgets, observability, capacity planning, incident management and post-incident analysis.
• You are strong in Python and have practical Rust experience, or you are willing to work deeply in Rust-based services.
• You have hands-on work with cloud platforms, Docker, Kubernetes and production service operations.
• You have built and maintained CI/CD pipelines, ideally on GitLab, and worked with infrastructure as code.
• You know distributed systems, microservices, APIs, networking and production debugging well.
• You can move quickly in a fast-changing AI ecosystem while keeping the rigor and risk awareness an SRE needs.
• You communicate clearly and collaborate across engineering, data, ML, security and product teams.

Conditions

The office is in Saint Mandé, reachable by metro line 1 (Saint Mandé station).

Benefits

• Profit sharing and a yearly company savings plan.
• 25 paid vacation days plus 12 additional paid days off.
• The company pays 50% of your transport pass.
• Lunch vouchers worth 9€ per day.
• Healthcare coverage for you and your family.
• Maternity leave of 20 weeks and paternity or co-parental leave of 7 weeks.
• A gym in the building.