Principal AI Engineer

RedRock Resourcing; Reading; On-site; GBP 85,000 - 125,000; Full time

Job summary

RedRock Resourcing is seeking three experienced Principal Software Engineers to join a brand-new team focussed on agentic engineering. You will work at the intersection of software engineering, architecture and AI to modernise complex legacy software, creating systems, standards and guardrails for safe, high-quality delivery.

This role is not a traditional ticket-driven position. You will be hands-on with complex legacy tech, using AI tooling to tackle debt and deliver value.

Qualifications

• Requires a seriously strong software engineering foundation.
• Deep experience with C#/.NET and large, complex or legacy systems.
• Comfort with AI-assisted development and modern tooling.

Responsibilities

• Use AI coding agents to investigate, refactor and modernise legacy codebases.
• Design the engineering harness that gives AI agents the right context, rules and constraints.
• Build repository-level instructions, agent skills, prompts and dynamic context.
• Work with technologies such as Claude Code, MCP and other agentic development tools.
• Establish automated testing and AI evaluation frameworks to balance speed and quality.
• Build guardrails, sandboxes and CI/CD controls around AI-driven development.

Skills

C#/.NET

Software architecture

Legacy systems experience

Tools

Claude Code

Cursor

GitHub Copilot

Codex

Job description

Principal Software Engineer - Agentic Engineering

We're looking for three experienced Software Engineers to join a brand-new team.

Using AI and agentic engineering to modernise complex legacy software, and fundamentally change how engineering teams build software.

This isn't a traditional software engineering role.

You won't simply be picking up tickets, writing code and moving onto the next sprint.

You'll be working at the intersection of software engineering, architecture, AI, developer experience and legacy modernisation, helping to design the systems, standards and guardrails that allow engineers to deliver better software safely.

What will you actually be doing?

You'll be hands-on with some seriously complex legacy technology, using modern AI tooling and engineering practices to tackle technical debt that has often been difficult to address through traditional approaches.

• Use AI coding agents to investigate, refactor and modernise legacy codebases
• Design the "engineering harness" that gives AI agents the right context, rules and constraints to work effectively
• Build repository-level instructions, agent skills, prompts and dynamic context
• Work with technologies such as Claude Code, MCP and other agentic development tools
• Establish automated testing and AI evaluation frameworks (evals) to make sure speed doesn't come at the expense of quality
• Build guardrails, sandboxes and CI/CD controls around AI-driven development

What background are we looking for?

We're looking for someone who has built a seriously strong software engineering foundation.

You'll ideally have deep experience in C#/.NET, alongside experience working with large, complex or legacy systems.

We're particularly interested if you've also been working with:

• Claude Code, Cursor, GitHub Copilot, Codex or similar coding agents
• Agentic development or AI-assisted software engineering
• Automated testing, evals and quality frameworks
• Legacy modernisation, major refactoring or framework migrations

You'll be joining during a major transformation across Technology, with significant investment in:

• Public cloud migration
• AI-enabled software engineering