Staff Software Engineer: Agentic AI Context & Memory

Scale AI; Greater London; On-site; GBP 120,000 - 160,000; Full time

Job summary

Scale AI in London, UK, is seeking a Staff Software Engineer to own end-to-end context and memory systems, working across embedding, vector store, and retrieval technologies. The role blends distributed systems engineering with applied ML, partnering with SWE and ML engineers to shape the platform used by enterprise and government customers.

You will design core primitives for memory, oversee evaluation methodologies, and set coding and architectural standards that endure beyond individual

Qualifications

• 8+ years of engineering experience, with production systems end to end.
• Experience with embeddings, vector search, retrieval augmented generation, fine-tuning, or agent memory architectures.
• Proficiency in Python and production ML infrastructure (containers, cloud, CI/CD).
• Experience setting technical standards and mentoring engineers.

Responsibilities

• Own large, ambiguous context and memory problems end to end from design to production.
• Architect primitives for retrieval, storage, and reasoning over long-running context.
• Design and maintain evaluation rubrics and benchmarks for memory and retrieval quality.
• Set technical standards and patterns adopted by the team.
• Collaborate with SWE and ML peers across memory and context workstreams.
• Debug production issues related to context retrieval and memory.

Skills

Python

Distributed systems

Embeddings

Vector search

Memory architectures

Tools

Containers

CI/CD

Cloud platforms

Job description

Scale AI in London, UK, is seeking a Staff Software Engineer to own end-to-end context and memory systems, working across embedding, vector store, and retrieval technologies. The role blends distributed systems engineering with applied ML, partnering with SWE and ML engineers to shape the platform used by enterprise and government customers.

You will design core primitives for memory, oversee evaluation methodologies, and set coding and architectural standards that endure beyond individual