Staff AI Research Engineer — Multi-Agent Systems

Seldon; Greater London; Hybrid; GBP 135,000 - 187,000; Full time

Benefits offered by this job

Equity

Job summary

Seldon is seeking an experienced Staff Engineer to join the Research Tribe in hybrid London, shaping core agent systems for production. You’ll design and implement scalable multi-agent architectures, with robust orchestration, memory models, and tool registries, working closely with researchers and engineers.

This role blends distributed systems, ML, and product impact. You’ll own behavior modeling, evaluation, prompt design, and production reliability, delivering measurable improvements and

Qualifications

• 5+ years building complex software and 2+ years shipping agentic LLM systems.
• Experience with multi-agent orchestration, sandboxed execution, and backends.
• Proficient in Python with async programming and structured data modeling.
• Comfortable shipping quickly and refining systems with tight feedback loops.

Responsibilities

• Design and implement agentic systems at scale with robust orchestration.
• Build agent governance: topology, memory, tool-use, and execution.
• Develop production-ready infrastructure for agents with tracing and cost monitoring.
• Create prompt schemas and reasoning templates integrated with systems.
• Transform raw agent outputs into high-signal insights.
• Ensure measurable, inspectable, and reliable agent behavior.

Skills

Multi-agent systems

Distributed systems

Python

Orchestration

Agentic LLMs

Production reliability

Tools

Pydantic

Async programming

Job description

Seldon is seeking an experienced Staff Engineer to join the Research Tribe in hybrid London, shaping core agent systems for production. You’ll design and implement scalable multi-agent architectures, with robust orchestration, memory models, and tool registries, working closely with researchers and engineers.

This role blends distributed systems, ML, and product impact. You’ll own behavior modeling, evaluation, prompt design, and production reliability, delivering measurable improvements and