Software Engineer

Agentic Software Engineer | West London | Full-time | Hybrid (4 days onsite per week)

Tech: AWS, Salesforce, Snowflake, Node.js, TypeScript, Python, React, GraphQL, Claude Code, Cursor, Copilot, Codex or similar agentic tools

We're hiring on behalf of a growing EdTech customer, a globally active education organisation with a genuinely AI-Driven engineering culture. Their tech team doesn't sit on the side of the business. It's central to it. How they build, what they automate, and how they adopt AI feeds directly back into what they teach. It attracts engineers who care about more than shipping tickets.

The role itself is one of the more forward-thinking briefs we've seen.

You won't be writing code. You'll be directing the agents that do, building the scaffolding that lets them run unsupervised, designing the validation systems that tell you output is good without reviewing every line, and enabling multiple agents to work in parallel across real enterprise problems.

This isn't about writing faster with Copilot. That's a speed boost. This is the next step, and they want someone already living on the other side of it.

What the role involves:

• Building agent scaffolding; context files, prompt architectures and skills that keep AI on track without constant intervention
• Designing verification systems for AI-generated code you didn't write or review
• Running parallel agent workstreams converging on shippable, tested output
• Working across a real enterprise stack: AWS, Salesforce, Snowflake, Node.js, TypeScript, Python, React, GraphQL
• Translating between stakeholders and agents; turning ideas into working product fast enough that iteration feels like a conversation

What they're looking for:

• 5+ years of software engineering experience — but your recent work looks fundamentally different
• You've used Claude Code, Cursor, Codex or similar agentic tools in production, not just demos
• You can write a spec an agent can execute without hand-holding
• You've thought seriously about how you verify AI-generated output you didn't produce yourself
• Strong communication across technical and non-technical stakeholders

What would make you stand out:

• You've built MCP servers, custom agent tooling, or agent skills
• You've implemented cost controls for token spend
• You can point to something non-trivial built primarily by agents under your direction

Please apply now to learn more!