Senior Software Engineer

WeDo; Reigate; Hybrid; GBP 70,000 - 100,000; Full time

Job summary

WeDo in Surrey is seeking a strong Full Stack Engineer who has already adopted an AI-first, agentic approach to developing software. This isn't a Copilot-type shortcut; it's about defining specs, prompts, and overseeing agent output while remaining technically accountable.

You'll join an established engineering team on a workflow automation platform used by 100+ clients. The team is moving toward an Agentic Software Development Lifecycle and will be the first within the business to fully adopt

Qualifications

• Proven experience with modern frontend frameworks, preferably Angular/TypeScript.
• Strong understanding of software architecture and design patterns.
• Hands-on experience applying AI-first and agentic development practices.

Responsibilities

• Define specifications and prompts for agents and review their outputs.
• Delegate engineering tasks to agents and oversee progress and quality.
• Develop reusable agent skills, workflows or development patterns.

Skills

Angular/TypeScript

System design

AI-first

Agentic development

Prompts for agents

Overseeing AI output

Agent skills

Job description

Surrey – Hybrid, 2 days per week in the office

Work Type:

Permanent

The Role:

We’re looking for a strong Full Stack Engineer who has already adopted an AI-first, agentic approach to developing software.

That doesn't mean using Copilot to write code faster.

We want someone who is already changing how they engineer software because of AI – defining specifications and prompts, delegating engineering tasks to agents, creating reusable agent skills and workflows, reviewing what comes back and remaining technically accountable for the end result.

You’ll join an established engineering team working on a workflow automation platform used by 100+ clients.

The team is now moving towards an Agentic Software Development Lifecycle, and they'll be the first engineering team within the wider business to properly adopt this way of working.

They need someone who's already started that journey and can help take the team with them.

What we're looking for:

• Full Stack experience, ideally with Angular/TypeScript, although other modern frontend frameworks are absolutely relevant.
• Strong understanding of system design, architecture and design patterns.
• Commercial experience using AI-first and agentic development practices as part of real software delivery.
• Experience defining specifications and prompts for agents rather than simply prompting an AI coding assistant for snippets of code.
• Experience delegating engineering tasks to agents and then overseeing, reviewing and challenging their output.
• Experience creating reusable agent skills, workflows or development patterns.
• Evidence that AI is being used across more of your software development lifecycle – development, testing, PRs, code review, QA, documentation or similar.
• Strong enough technically to challenge AI-generated code and decisions around architecture, security, quality, maintainability and system design.
• Someone who understands when to trust an agent, when to redirect it and when to override it completely.
• Someone who can bring what they've learned to an established engineering team and help other engineers adopt a more agentic way of working.

Why is this interesting?

This isn't an AI Engineer role.

And it's not a traditional Full Stack role with "AI" added to the job title.

They're looking for a Software Engineer who believes the way we build software is fundamentally changing and has already started adapting to that change.

The product itself has an evolving AI roadmap covering natural language workflow creation, agentic capabilities, MCP and LLM integration, but they don't want to over-index on experience building AI products.

The bigger opportunity is changing how the team builds software.

You’ll be joining the first team within the business to properly adopt an agentic development approach and helping answer some genuinely interesting engineering questions:

How much of the SDLC can we delegate to agents? How should we specify the work? What reusable agent capabilities should we build? Where does human judgement need to remain? And how do we make sure engineers are still technically accountable for what gets shipped?

If your AI experience amounts to "I use Copilot every day", this probably isn't the role.

If you've genuinely started building software differently because of agents and can talk specifically about how you've used them, what you've delegated to them and how you've managed and challenged their output, we'd like to speak with you.