Tech Stack
AzureCloudETLITSMReactServiceNowVault
About the role
- Own the technical delivery of a production enterprise AI agent platform for a large organisation
- Lead 12-15 engineers across multiple teams and drive architectural decisions
- Ship working agents to production within weeks
- Navigate Azure's quirks whilst building on LangGraph/LangChain and manage model deployments
- Demo to C-level execs and communicate risk to avoid hallucination or security concerns
- Implement multi-agent orchestration using ReAct patterns and Model Context Protocol
- Build RAG pipelines with hybrid search, vector indexing and semantic search
- Integrate with enterprise systems (SharePoint, ServiceNow, Microsoft Teams/Entra ID) and implement zero-trust private endpoints
- Set up observability, ETL pipelines, secrets management, rate limiting and security policies
- Turn academic papers on agentic AI into production-grade working code
Requirements
- Lead 12-15 engineers across multiple teams
- Built production LLM systems handling real workloads
- Led distributed teams through technical ambiguity
- Shipped when perfect wasn't possible (and can explain the trade-offs)
- Deep Azure experience (Azure Container Apps, Azure AI Search, Cosmos DB, Azure API Management, Azure AI Foundry, Azure Functions, Azure Key Vault, Azure Application Insights, Azure Blob Storage)
- Experience with LangGraph/LangChain and ReAct pattern orchestration
- Knowledge of Model Context Protocol (MCP) for tool integration
- Experience with LiteLLM/Portkey for model gateway routing
- Experience building RAG pipelines with hybrid (vector + keyword) search
- Experience with Langfuse or other LLM observability tools
- Familiarity with multiple LLM providers (Azure OpenAI, Anthropic, open models)
- Experience with enterprise integrations (SharePoint via Graph API, ServiceNow, Microsoft Teams/Entra ID, CIAM)
- Experience with enterprise authentication (OAuth, SAML, JWT tokens)
- Opinions on vector databases and ability to defend choices in architecture review
- Comfort presenting to both engineers and C-level executives
- Ability to turn academic papers on agentic AI into working code
- Experience shipping production systems (not PoC)
- Experience navigating Azure-specific constraints when building agent platforms