AI Platform Engineer

About the role

The IRC is investing in enterprise AI to improve the speed, quality and scale of its programmes and operations, using platforms such as Anthropic, OpenAI and integrated data environments. The IT department is building technical capacity to keep these systems running reliably, securely and effectively. The AI Platform Engineer owns the technical layer behind the AI tools: systems, integrations and infrastructure. This is a technical operations role, not user-facing, working closely with the AI Platform Support Director.

Key responsibilities

• Administer the IRC's enterprise AI environments (currently Anthropic and OpenAI), manage user access, API keys, workspace settings and monitor health, usage and API performance
• Improve observability (uptime, error rates, token consumption, integration reliability) and oversee documented configuration changes and deployments
• Flag anomalous usage and coordinate with InfoSec on responsible use and access controls
• Work with DevOps, software and data engineering teams on deployment processes and infrastructure dependencies
• Follow change management procedures, including documentation, testing and approvals before production
• Build lightweight scripts, connectors and automations in Python and/or JavaScript/TypeScript, and troubleshoot integration failures and API connectivity issues
• Collaborate with data engineering on AI/KM pipelines, including vector store ingestion, retrieval configuration and source data connections
• Track AI resource utilisation, report on cost efficiency, and act as technical point of contact with InfoSec on security, data handling and compliance
• Support risk assessments, help develop technical SOPs and best practices, and act as a technical resource for programme and operations teams
• Take part in rollout planning, onboarding documentation and sprint/project planning

Required profile

• Hands-on experience administering enterprise AI platforms (Anthropic, OpenAI, Azure OpenAI or comparable)
• Familiarity with LLM infrastructure: prompt pipelines, MCP, tool-calling frameworks, vector databases, RAG patterns and embeddings
• Experience with Databricks or similar data/ML platforms is a strong plus
• Ability to read and work in existing codebases, with version control (Git) and deployment practices
• Experience monitoring distributed systems or SaaS platforms, including alerts, logs and performance diagnosis
• Working knowledge of information security principles for SaaS and APIs, and ability to engage with InfoSec teams
• Clear communication of technical concepts to non-technical colleagues and strong documentation habits