Backend Software Engineer (MCP & AI Agents)

We are seeking a skilled Backend Software Engineer to build and maintain production-ready Model Context Protocol (MCP) servers that power our AI agent workflows. This role focuses on the Internal AI platform, requiring you to design distributed systems using Python, FastMCP, and Redis with enterprise-grade security.

About The Role

You will join a team responsible for the backbone of clients ai infrastructure. You will be responsible for designing and deploying FastMCP servers that handle complex agentic workflows. Your work will center on creating robust, asynchronous systems that integrate securely with Azure AD and persist state via Redis. This role demands high code quality, rigorous testing habits (TDD), and close collaboration with platform and delivery teams to ensure scalable, secure AI solutions.

Requirements:

·       Core Development: Strong proficiency in Python with deep knowledge of async/await patterns.

·       Solid experience with FastAPI and REST API design.

·       MCP Specialisation: Proven experience building MCP servers (critical requirement).

·       Hands-on experience with FastMCP framework.

·       Data & Storage: Proficiency with Relational Databases (SQL) and Key-Value stores (Redis).

·       Security (IAM): Deep understanding of AuthN/AuthZ protocols. Experience with OAuth2, Azure Entra ID/SSO, and JWT token handling.

·       Architecture: Familiarity with Distributed Systems and Event-Driven Architecture. Experience with Docker and containerization principles.

·       Soft Skills: Strong problem-solving capabilities. Ability to work independently and communicate technical concepts clearly.

Nice to Have

• MCP Ecosystem: Orientation in the broader ecosystem of MCP applications and clients

• AI Frameworks: Experience with AI/LLM agent frameworks such as LangGraph, LangChain or Langfuse.

• DevOps: Familiarity with Azure CI/CD pipelines and Github Actions.

• Observability: Knowledge of Datadog or similar platforms for logging and monitoring.

• Modern Practices: Prior exposure to “Vibe” coding practices (AI-assisted iterative coding).