Staff Engineer, Scalable Agent Runtimes & Systems

Callosum; Greater London; On-site; GBP 120,000 - 180,000; Full time

Benefits offered by this job

Competitive Salary

Equity & Ownership

Private healthcare

We offer Visa sponsorship and Reloc."

London office – in person work

Job summary

Callosum, the Intelligent Systems Company, is building an agent runtime for long-running workflows and distributed infrastructures in London. You will own the runtime and programming model for durable, scalable agent systems, focusing on execution, checkpointing, and observability across the stack.

We expect strong expertise in distributed systems, Python, and a systems language like Rust/C++/Go, plus experience with modern ML tooling and open‑source projects.

Qualifications

• Strong background in distributed systems and durable workflows.
• Proficient Python programming and systems language proficiency (Rust/C++/Go).
• Experience designing libraries, runtimes, DSLs, or workflow engines that are adopted by engineers or researchers.

Responsibilities

• Design and build a stateful, asynchronous agent runtime across distributed environments.
• Define event histories, checkpoint/resume semantics, deterministic reconstruction, retries, timeouts, and recovery from partial failure.
• Develop traces, replay, and time-travel debugging across model calls, tools, and runtime decisions.
• Create clear Python APIs, libraries, and DSLs with stable extension points for tools and backends.
• Implement scheduling, placement, resource accounting, sandboxing, and isolation for secure tool execution at scale.
• Develop tooling for trajectory collection, performance profiling, experiment comparison, and regression detection.

Skills

Distributed systems

Python

Rust/C++/Go

Debugging

API design

Tools

Ray

Temporal

Kafka

Erlang/BEAM

PyTorch

JAX

TensorFlow

Job description

Callosum, the Intelligent Systems Company, is building an agent runtime for long-running workflows and distributed infrastructures in London. You will own the runtime and programming model for durable, scalable agent systems, focusing on execution, checkpointing, and observability across the stack.

We expect strong expertise in distributed systems, Python, and a systems language like Rust/C++/Go, plus experience with modern ML tooling and open‑source projects.