Graduate AI Automation Engineer

Byrne Wallace Shields is building out its internal AI automation capability, and we're looking for a graduate engineer to help us do it. Working directly with our Head of Digital Automation with AI, you'll design, build and support the tools, integrations and agent workflows that automate how the firm runs — connecting AI models such as Claude and ChatGPT to our Microsoft 365 estate, building the APIs and MCP servers that let those models work safely with firm data, and helping shape the engineering practices a growing AI function needs. It's a hands-on role and a genuine chance to help build a new function from the ground up, with broad exposure across the full stack of AI automation work.

What you'll be doing

• Build, test and maintain internal tools and automations that use large language models (Claude, ChatGPT, Microsoft Copilot) to take manual effort out of firm processes.
• Design and develop APIs and integrations — including MCP (Model Context Protocol) servers — connecting AI models to systems such as SharePoint, Dataverse and Microsoft Graph securely and auditably.
• Write clean, well-tested code in Rust and Python (with UV), and TypeScript/Node where it fits; work in GitHub day to day — branching, pull requests, code review and Actions CI/CD.
• Contribute to prompt design, agent orchestration and evaluation, measuring whether a workflow is accurate, reliable and safe enough to put in front of fee-earners.
• Support Power Platform solutions (Power Automate, Power Apps, Dataverse) where they fit, document what you build, and work with legal and business services teams to gather requirements and demo solutions.
• Apply the firm's AI governance, data protection and information security requirements throughout, and stay current with a fast-moving field.

What we're looking for

• AI tooling: practical, hands-on experience with Claude and ChatGPT — their APIs and developer tooling (e.g. Claude Code or the OpenAI SDKs), not just the chat interfaces; you understand prompting, tool use, context management and model limits.
• GitHub: confident with Git and GitHub (branching, pull requests, code review, Actions); you can point us at repositories you've contributed to.
• Rust & Python: working knowledge of Rust with at least one non-trivial project you can walk us through; solid Python fundamentals with UV.
• APIs: designing, building and consuming REST APIs — auth (OAuth 2.0 / bearer tokens), JSON, error handling, rate limiting.
• Fundamentals: you write tests and readable code, debug methodically, explain technical work to non-technical colleagues, and write documentation people actually read.
• A degree (or equivalent) in computer science, software engineering, data science or a related discipline — or a demonstrable portfolio of equivalent work. Curious, self-directed, and serious about quality and confidentiality.