AI Engineer

UK, hybrid (fully remote within Europe considered for a 100% fit)  ·  Full-time, 12 weeks initial, extension up to 6 months.

A B2B SaaS platform is kicking off a rapid AI enablement programme. They need a hands-on senior lead to take identified automations from idea to production, measured, governed and actually adopted by teams. The role sits within product engineering. It starts with Product, QA and Solutions delivery, then scales into a reusable orchestration framework across the business.

This is a high-ownership role: one named lead, personally accountable for delivery, stakeholder alignment and handover.

What you’ll do

- Assess and push forward a few in-flight automations, then deliver a portfolio of agentic workflows: QA & test generation, bug triage, requirements, solution implementation, docs, prototyping

- Design the target architecture: how people, agents and existing systems share work, using reusable integration/orchestration patterns

- Build in the guardrails: evals & thresholds, human approvals, audit trails, monitoring, failure handling

- Set baselines and measure real impact (not demo metrics)

- Train internal owners and hand over maintainable, documented workflows

Must have

- Personally designed and shipped production agentic / multi-step LLM workflows in a SaaS or software company, with measurable outcomes

- Strong hands-on Python or TypeScript + AWS in production; integrations across APIs, databases, events, queues

- Evals, observability, access control, human-in-the-loop design; MCP & secure tool integration

- Ability to select and govern AI models/tools (data handling, retention, security, limitations)

- AI-assisted SDLC with proper quality controls

- Credible with both executives and engineering teams

Nice to have

Laravel/PHP (existing platform) · multi-tenant SaaS · Amazon Bedrock / AgentCore · Google ADK, Strands or similar · AWS serverless (Lambda, SQS, Step Functions, EventBridge) · GDPR, ISO 27001, SOC 2, ISO 42001 environments

Probably not a fit if your AI experience is POCs/demos only, advisory without hands-on coding, or primarily ML research/model training.

Be ready to show: real production examples with before/after results, an architecture you personally designed, one production failure and how you fixed it, and a reference.