Agentic Staff Engineer (AI)

TL;DR

Agentic Staff Engineer (AI): Design and operate data pipelines, datasets, and evaluation workflows for reliable AI systems with an accent on agentic patterns, platform reliability, and governance. Focus on owning technical vision for multi-step agent workflows, evolving orchestration stack, and leading cross-team initiatives for safety and compliance.

Location: Hybrid in Paris area (Issy-les-Moulineaux, France). Professional English proficiency required (written and spoken).

Company

Video platform with AI-powered products for user engagement, professional hosting, and advertising, with 400 people across France, New York, and Singapore.

What you will do

• Own technical vision and architecture for agentic systems including multi-step workflows, tools, memory, planning, and human-in-the-loop controls.
• Design and evolve internal agent orchestration stack: tool registry, permissions, execution lifecycle, state management, and audit trails.
• Define standards for agent reliability: traceability, replay, evaluation harnesses, fallbacks, and degradation modes.
• Lead cross-team initiatives with Product, Data, Infrastructure, Security/Privacy, and ML to align stakeholders and ensure delivery.
• Set direction on platform choices like model routing, caching, prompt management, observability, and governance.
• Drive adoption via libraries, documentation, enablement, and mentor senior engineers through reviews and operational excellence.

Requirements

• Professional English proficiency required (written and spoken).
• Extensive experience designing and operating large-scale distributed systems (reliability, performance, cost, security).
• Deep understanding of agentic patterns and failure modes (tool calling, orchestration, memory, planning, evaluation).
• Proven technical leadership across teams: setting standards, influencing roadmaps, delivering platforms.
• Strong system design skills: multi-tenant architectures, permissions, data governance, safe execution.
• Strong operational mindset: observability, debugging, incident response, corrective actions.

Nice to have

• Experience building evaluation systems at scale (test suites, synthetic data, human review, A/B testing).
• Understanding of privacy-by-design and compliance for user/content data (PII, retention, access).
• Experience with retrieval systems (vector search, hybrid, ranking) and LLM cost-control.
• Familiarity with event-driven patterns (queues, retries, idempotency) and platform engineering.

Culture & Benefits

• Hybrid work framework with ability to work 3 months abroad.
• High-end health insurance, meal vouchers, public transport/bike refund, fitness subscription.
• Extended paternity/coparental leave, paid time off, RTT, saving time plan.
• Career development: training, internal mobility, compensation cycles, 360 feedback.
• Living employee culture: events, trainings, parties, all-hands.
• Saving Plan Canal+, personal services vouchers, European Economic and Social Committee perks.