Backend Software Engineer — Agentic AI Systems

Dentsu is an integrated growth and transformation partner specializing in marketing, technology, and consulting. They are seeking a Backend Software Engineer to build and own production-grade Python APIs and architect orchestration pipelines for their AI & Innovation team, focusing on developing an enterprise-level agentic AI platform.

Responsibilities:

• Build and own production-grade Python APIs that power multi-agent workflows
• Architect orchestration pipelines using tools like Airflow, Dagster, or Prefect—handling retries, fallbacks, and long-running processes
• Solve real rate limiting and model routing challenges across LLM providers (OpenAI, Anthropic, and others)
• Design error handling and tracing systems so failures are caught, routed, and fixed proactively
• Set the standard for testing and reliability in a codebase moving from demo to production

Requirements:

• 2–5+ years of backend Python experience building enterprise-grade, production systems
• Designed systems with concurrent processing and know how to handle scale failures
• Worked with workflow orchestration tools and understand the tradeoffs
• Familiar with LLMs in production—their APIs, their limits, and how to build around them
• Balance pragmatism with architecture—knowing when not to over-engineer
• Want to own something from the ground up, rather than inherit a legacy codebase