Summer 2027 Graduate Intern

About the Role

The Federal Reserve Board (Board) provides students opportunities for career success through its Internship program, designed to offer valuable work experience for undergraduate and graduate students. This role is within the Application Infrastructure Support (AIS) unit of the Division of Information Technology, focusing on the Board's governed platform for connecting SI/AI tools to enterprise data and systems. The platform designs, hosts, and operates Model Context Protocol (MCP) servers, retrieval-augmented generation (RAG) services, and an identity, gateway, and inventory layer for secure SI/AI application access. The platform runs on Kubernetes (Amazon EKS) within the Board's AWS environment, built and delivered using infrastructure as code and CI/CD pipelines.

Key Responsibilities

• Build and deploy MCP servers that expose Board data sources and tools to SI/AI clients, using Python and the MCP SDK.
• Set up and test RAG components: document ingestion, chunking, embeddings, and vector search on PostgreSQL with pgvector.
• Deliver platform changes using Terraform for cloud infrastructure, Ansible for configuration, GitLab CI pipelines for build and test, and ArgoCD GitOps for deployment to Kubernetes.
• Contribute to the identity and access layer for SI/AI connectors, including OAuth client registration, audience-bound tokens, and routing through the LiteLLM gateway.
• Populate and maintain the MCP connector inventory and catalog.
• Write templates, documentation, and quick-start guides for building connectors to enterprise standards.
• Participate in daily stand-ups, sprint planning, and design reviews.

Requirements & Tech Stack

• Python
• Kubernetes (deployments, services, Helm or Kustomize)
• AWS (IAM, EKS, Secrets Manager)
• Terraform
• Ansible
• GitLab CI (or equivalent pipeline tool)
• ArgoCD (or other GitOps tools)
• PostgreSQL with pgvector
• Model Context Protocol (MCP)
• Retrieval-Augmented Generation (RAG)
• Linux
• Git
• Docker (or equivalent container technology)
• REST APIs, JSON, relational databases, and SQL
• OAuth 2.0 and OpenID Connect concepts (exposure to Keycloak or Okta is a plus)
• SI/AI coding assistants (Claude Code, Copilot, or similar)

Qualifications

• Currently enrolled in an undergraduate or graduate degree program at an accredited university and returning to continue studies after the internship.
• Graduating August 2027 or later.
• Minimum 3.0 GPA.
• US Citizenship required.
• Programming in Python.
• Working knowledge of large language model tooling and concepts: prompting, tool calling, embeddings, and retrieval-augmented generation.
• Familiarity with infrastructure as code and CI/CD practices.
• Working knowledge of Linux, Git, and containers.
• Understanding of REST APIs, JSON, relational databases, and SQL.
• Ability to write clear technical documentation.

Preferred Qualifications:

• Hands-on experience with the Model Context Protocol or building tools for SI/AI agents.
• Experience using SI/AI coding assistants and validating their output.