In this role you will shape Multiverse into an AI-first company by building end-to-end agent systems that integrate with the current product. You will work across engineering to redesign how the company operates, improve learner outcomes, and set the standard for AI-driven product development. You will balance context management, tooling, and evaluation to deliver scalable, trusted AI solutions. This is a chance to influence industry best practices and help Megaverse scale its AI-enabled education platform.
Pay / Benefits
27 days holiday plus 5 wellbeing days
private medical insurance with Bupa
life insurance
gym membership & wellness resources
Wellhub and Spill mental health support
hybrid work (London) with three days in office (location varies)
Responsibilities
Own and deliver complete agent systems—architecture, implementation, evaluation, and production operation
Design context and retrieval strategies for context windows, including memory, summarisation, and chunking
Build evaluation frameworks with automated pipelines and human-in-the-loop processes
Design tool integrations (APIs, data contracts, error handling) for reliable agent interactions
Influence technical direction through architecture and decision-making
Raise the bar through code review and pair programming to raise production-quality AI engineering
Use Claude Code as the primary development workflow and mentor others in AI-assisted development
Key requirements
Shipped AI systems at scale serving real users
Deep understanding of context management, retrieval, chunking, and summarisation
Experience with model selection, routing, and cost engineering
Proficient in tool use and agent augmentation with external systems
Experience building evaluation frameworks for AI outputs
Full-stack delivery capabilities across LLM integration, backend services, data pipelines, and some frontend
Strong communication skills for documenting designs and giving feedback
Strong communication
Product thinking and user-centric problem solving
Collaborative and cross-functional teamwork
Context management and retrieval strategies
Chunking, conversation memory, summarisation
Model selection, routing, and cost engineering (token economics, caching, batching)
We use essential cookies to make this site work. With your permission we would also like to use the optional cookies below. Nothing optional is set until you choose, and you can change your mind at any time from "Cookie settings" at the bottom of every page. Cookie policy.
Essential — keep the site working and secure, and remember the choice you make here. Always on.