AI Developer

AI Solutions Engineer

Location: Bevis Marks, London – near Liverpool Street

Hours: Monday to Friday, normally 09:00–17:30

Working pattern: Office based initially, with up to 2 days working from home once fully up to speed

About the role

We are looking for an AI Solutions Engineer to join our London-based MSP team and help our clients put AI to work properly.

This is a hands-on role for someone who understands how businesses actually operate and can build the technology to improve it. You will sit with clients, map their workflows, and then build the AI agents, automations and small apps that make a real difference.

We want someone who understands code and security as well as business process. Everything you build will run on client data, so it needs to be secure and well documented, not just clever.

You will spend the majority of your time on client work, with part of the role focused on improving our own internal AI and automation tools.

We are a friendly and supportive team and want someone who enjoys solving problems, learning quickly and seeing things through to a working result.

What you will be doing

• Run discovery workshops with clients to understand how their teams work and where AI or automation can help.
• Map current workflows and design better ones that clients can review and sign off.
• Capture requirements, data sources and success measures before any build starts.
• Build AI agents and assistants, mainly on Microsoft 365, Copilot Studio and Power Platform.
• Build workflow automations that connect business systems through connectors and APIs.
• Write code in Python, TypeScript or C# where low-code tools are not enough.
• Use AI coding tools to prototype small apps and dashboards, then review and test everything they produce.
• Design every solution around least-privilege access, and protect against prompt injection and data leakage.
• Help clients with AI readiness, governance and practical AI usage policies.
• Run pilots, train users and measure results against what was agreed.
• Hand solutions over to the Service Desk with clear documentation and runbooks.

Internal AI tooling

Part of the role, agreed each quarter, will involve improving how Cubit works, including:

• Automating Service Desk triage, documentation and reporting.
• Spotting repetitive internal processes and proposing automations.
• Building reusable templates and patterns so each client project is quicker than the last.
• Helping shape our AI services as the practice grows.

This is not a pure development role. The majority of your time will be spent working with clients to understand and solve real business problems.

What we are looking for

We are looking for someone with at least 3 years’ experience in a role that mixed building technical solutions with working directly with business users.

You should be confident with areas such as:

• Building and deploying AI agents or LLM-based automations.
• Microsoft 365 and Power Platform.
• Copilot Studio or similar agent platforms.
• Python, TypeScript or C#.
• REST APIs, webhooks and authentication.
• Entra ID, identity and permissions.
• AI security risks such as prompt injection and data leakage.
• Workflow mapping and requirements gathering.
• Using AI coding tools productively.

You should also be:

• Comfortable running a workshop and asking the right questions.
• Able to explain technical choices clearly to non-technical clients.
• Disciplined about testing, documentation and security.
• Practical, preferring to ship something useful and improve it.
• Curious and quick to pick up new tools.
• Organised and able to manage several client projects at once.

Nice to have:

• Microsoft certifications such as PL-200, PL-400 or AI-102.
• MSP or IT consultancy experience.
• Experience with open-source workflow automation tools or MCP.
• Knowledge of the OWASP Top 10 for LLM Applications, Cyber Essentials or ISO 27001.

Working pattern

Normal working hours are 09:00–17:30, Monday to Friday.

The role is office based while you learn our clients, systems and processes. Once fully up to speed, there is the option to work from home for up to 2 days per week. Discovery workshops and pilots will sometimes be run at client offices across London.

Benefits

• Up to 2 days working from home once fully up to speed.
• Perkbox membership.
• Private health scheme.
• Quarterly team events.
• Microsoft certifications and training
• Budget for AI tools -…