ML Research Engineer - Member of Technical Staff

About the role

Callosum is building an orchestration layer that unifies heterogeneous compute across the full stack to enable the next generation of AI systems. As an ML Research Engineer, you will develop evidence-based methods for building agentic systems at scale by designing experiments to isolate failure points and creating durable analysis infrastructure. You will have the latitude to choose research problems that directly shape the company's future technical direction.

Responsibilities

• Design and run experiments to isolate failure points in agentic systems across behavior, traces, and activations.
• Develop agent analysis and evaluation infrastructure including observability, behavior analysis, and auditing tools.
• Research core themes such as context and memory management, agent steering, task decomposition, and continual learning.
• Publish research findings, methods, and discoveries as both internal assets and public evidence.
• Collaborate across teams to solve domain-specific application problems or invent novel model and agent architectures for unconventional silicon.

Required skills

• LLMs
• Agentic systems
• Experimental design
• Software engineering
• Research

Nice to have

• RL
• Post-training
• Multi-agent systems
• Planning
• Program synthesis
• SGLang
• vLLM
• Neuroscience
• Distributed systems

Qualifications

• Evidence of ability to conduct independent research and produce results
• Deep hands-on experience with LLMs in agentic settings
• Strong engineering capability in shared codebases

Benefits

• Competitive salary
• Equity
• Private healthcare
• Visa sponsorship
• Relocation benefits

About the Company

Callosum is an Intelligent Systems Company based in London that builds the software orchestration layer to unify heterogeneous compute across the full stack. They recently raised $100M to build infrastructure that co-evolves models, workflows, and silicon to deliver inference tailored to every workload.