AI Platform Engineer

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for an AI Platform Engineer based in United Kingdom. Join a forward-thinking technology team building the AI platform that powers intelligent agents across multiple business functions, including engineering, business development, compliance, support, and finance. In this hands-on individual contributor role, you will design, develop, and operate the infrastructure that enables AI agents to perform increasingly autonomous tasks safely and reliably. You will work across agent runtimes, integrations, identity management, security guardrails, evaluation frameworks, and observability. Collaborating closely with internal domain experts, you will turn operational challenges into scalable AI-powered solutions that deliver measurable business value. You will take ownership from initial design and implementation through production deployment, monitoring, and continuous improvement. This remote-first opportunity is ideal for an experienced engineer who combines strong software fundamentals with practical expertise in deploying and scaling LLM-based systems. Accountabilities You will help build and evolve the organization's AI platform, enabling teams across the business to deploy reliable, secure, and increasingly autonomous agents. Your responsibilities will include: Productionizing AI workflows: Transform prototypes into robust production systems with defined service-level agreements (SLAs), supporting use cases such as specification-to-pull-request agents, AI-powered code reviews, incident triage, and Know Your Business (KYB) reviews. Building and managing MCP servers: Design, deploy, secure, and maintain Model Context Protocol servers, including authentication, per-user OAuth, scoped access, and integrations with tools such as Slack, GitHub, Jira, HubSpot, and internal services. Developing autonomous agents: Build multi-step agents capable of planning and taking action across different tools, progressing through controlled trust stages from shadowing and suggesting to executing tasks autonomously. Implementing security and governance: Establish platform-wide guardrails, including agent identities, scoped credentials, audit logs, CI checks, approval workflows, allowlists, and least-privilege access policies. Creating evaluation frameworks: Develop test datasets, evaluation criteria, regression suites, and live production scoring systems to measure agent quality and connect performance metrics to business outcomes. Improving observability: Instrument agents and workflows with tracing, logging, latency and cost metrics, drift detection, alerts, and dashboards that help internal Operators monitor performance. Managing AI expenditure: Track usage and spending across AI providers, including Anthropic, OpenAI, and OpenRouter. Establish budgets, alerts, and cost attribution at the team, agent, and workflow levels. Supporting internal stakeholders: Work directly with Operators across business functions to understand their needs, develop requested features, create documentation, deliver enablement sessions, and build self-service tools. Developing reusable AI capabilities: Create and maintain shared skills, plugins, and knowledge resources covering company processes, standards, security requirements, and brand guidelines. Driving process automation: Proactively identify repetitive manual workflows that could benefit from AI, implement automation solutions, measure their impact, and continuously improve them. Taking end-to-end ownership: Write technical specifications, plan and deliver projects independently, communicate progress asynchronously, and remain accountable for monitoring and improving solutions after deployment.

Requirements

The ideal candidate combines strong software engineering fundamentals with hands-on experience building and operating production-ready AI systems. You should be comfortable working independently, communicating with technical and non-technical stakeholders, and applying sound judgment to complex automation and security challenges. Required qualifications include: At least 5 years of experience as a software, platform, or infrastructure engineer, with a solid understanding of system design, data modeling, testing, debugging, and code review. Between 6 and 12 months of experience shipping LLM-based systems to production and supporting agentic workflows used by real end users. Daily, hands-on experience using AI coding agents such as Claude Code and Codex to write, test, and review production code. Experience developing autonomous or multi-step agents with tool integrations, using technologies such as the Claude Agent SDK, Anthropic APIs, OpenAI APIs, or similar frameworks. Practical experience building MCP servers, including OAuth authentication flows, scoped tokens, and user-level access controls. Knowledge of identity and access management systems such as Okta or similar identity providers, including SSO protocols (SAML/OIDC), SCIM provisioning, group-based access, and device policies. Experience creating automated system identities, including GitHub Apps, service accounts, API keys, and managing secrets through tools such as 1Password. Understanding of security guardrails and governance practices, including least-privilege access, audit logging, CI checks, approval gates, allowlists, and policy-as-code. Experience designing evaluation systems for LLM applications, including test sets, evaluation rubrics, CI regression testing, release gates, and live production scoring. Practical experience implementing production observability through tracing, metrics, alerting, and dashboards, using tools such as OpenTelemetry, Grafana, Datadog, Langfuse, LangSmith, or Braintrust. Experience monitoring and controlling AI expenditure across providers through API key limits, cost centers, budgets, and actionable alerts. Strong programming skills in at least one of Typ