Jarvis is supporting a client who is adding AI gateway capability to its existing Apigee X production API gateway to govern LLM and MCP traffic. Apigee X is already in production but there is no AI gateway in place today, and the first release is to stand it up across all environments and route existing Azure AI Foundry model deployments and the Intelligent Document Processing (IDP) service through it.
****This is an initial 6 month contract term and is a hybrid role with 2 days/week on-site in downtown Toronto,
Target models are Azure AI Foundry (Azure OpenAI and Claude models), with Databricks Model Serving in a later phase. The senior engineer leads design and production delivery of the AI gateway, works hands-on end to end across Apigee and Azure, sets the reusable patterns the pod builds on, and guides the junior engineer. This is AI gateway engineering - LLM routing, token governance and AI policy configuration - rather than traditional REST API management.
What you will do
- Lead the design and stand-up of the AI gateway across all environments, then route all existing Azure Foundry model deployments and the IDP service through it (first release by end of Q1 2027).
- Build and configure Apigee AI gateway policies
- Configure the MCP gateway: route internal and external MCP servers through the gateway with authentication, authorization and tool-level access control; build MCP servers on top of existing application APIs where needed.
- Develop API proxies and integrate with Azure OpenAI; provision Azure AI Foundry models.
- Implement authentication and authorization with Entra ID and Key Vault, including OAuth for MCP.
- Write automation scripting and configure logging to Splunk and other Canada Life monitoring tools.
- Lead production go-live in Q1 2027, then own enhancements and production support from Q2 2027.
- Set reusable patterns and standards.
Must-have experience
- Hands-on Apigee (X / hybrid) policy and proxy development in production - traffic policies, API proxies, deployment across environments.
- Demonstrable AI gateway work: LLM routing, token and cost governance, and AI guardrails, as distinct from traditional REST API management.
- Azure OpenAI / Azure AI Foundry integration and model provisioning.
- MCP (Model Context Protocol): gateway routing, MCP server development on top of existing APIs, tool-level access control, and OAuth for MCP.
- Data protection in the request/response path: PII de-identification with re-identification (Sensitive Data Protection or equivalent) and prompt/response screening (Model Armor or equivalent).
- Identity and secrets: Entra ID, Key Vault, and OAuth.
- Automation scripting and observability/logging to Splunk or comparable monitoring tools.
- Proven design leadership: able to lead design and production delivery end to end.
Nice to have
- Databricks Model Serving.
- Experience integrating Claude / Anthropic models alongside Azure OpenAI.
- Semantic caching implementation at scale.
- Financial services or other regulated-industry delivery experience.
Apply today to learn more!