MCP Platform Engineer

Overview
We are seeking an experienced MCP Connector Developer to design, build, and maintain Model Context Protocol (MCP) server implementations that enable AI agents and LLMs to interact with enterprise systems, APIs, and data sources. This role sits at the intersection of AI/LLM integration and enterprise software engineering, requiring expertise in building robust, secure connectors that power intelligent agent workflows.

Key Responsibilities
- Design and develop MCP server implementations exposing tools, resources, and prompts for LLM consumption
- Build secure connectors integrating enterprise platforms (Splunk, Zendesk, Databricks, ServiceNow, etc.) with AI agent systems
- Implement JSON-RPC protocol handlers and manage client-server communication patterns
- Develop tool schemas and parameter validation logic for LLM tool calling
- Create error handling, retry mechanisms, and circuit breakers for production reliability
- Optimize connector performance for low-latency AI agent interactions
- Implement authentication flows (OAuth, API keys, SSO) and secure credential management
- Write comprehensive tool documentation including descriptions, parameters, and usage examples that LLMs can understand
- Build testing frameworks for connector reliability and LLM interaction validation
- Collaborate with AI/ML engineers to understand agent requirements and context limitations
- Monitor connector usage patterns and optimize for token efficiency
- Maintain and version existing MCP servers as APIs and LLM capabilities evolve

Required Qualifications
- 4+ years of software development experience with production systems
- Strong proficiency in Python and/or TypeScript/Node.js
- Deep experience with REST APIs, GraphQL, and various integration patterns
- Understanding of JSON-RPC, WebSocket, or similar RPC protocols
- Expertise in async/await patterns and concurrent programming
- Experience with authentication mechanisms (OAuth 2.0, JWT, API keys)
- Strong grasp of error handling, logging, and observability practices
- Experience building SDKs, libraries, or developer tools
- Knowledge of security best practices for API integrations and data handling
- Familiarity with Git, CI/CD pipelines, and deployment automation

Preferred Qualifications
- Hands-on experience with Model Context Protocol (MCP) specification and implementations
- Experience integrating with LLM APIs (OpenAI, Anthropic, Azure OpenAI, Google Vertex AI)
- Understanding of AI agent frameworks (FastMCP)
- Knowledge of prompt engineering and LLM tool calling mechanisms
- Experience with function calling and structured output from LLMs
- Familiarity with enterprise platforms (Splunk, Databricks, Zendesk, Salesforce, Jira)
- Understanding of token optimization and context window management
- Experience with schema validation (JSON Schema, Pydantic, Zod)
- Knowledge of containerization (Docker) and orchestration (Kubernetes)
- Background in observability tools (Prometheus, Grafana, Datadog)
- Contributions to open-source AI/LLM projects

Technical Skills
- Languages: Python 3.10+, JavaScript (Node.js 18+)
- Protocols: JSON-RPC 2.0, REST, GraphQL, Server-Sent Events (SSE), WebSockets
- LLM Integration: OpenAI API, Anthropic Claude API, Azure OpenAI, function calling, tool use
- Frameworks: FastAPI, Express.js, async/await patterns, Agent SDK integration
- Data: JSON Schema, Pydantic models, data validation and serialization
- Tools: Git, Docker, pytest, Jest, VS Code, Postman/Insomnia
- Security: OAuth 2.0, JWT, encryption (AES, RSA), secure secret management
- Concepts: API design, rate limiting, retry logic, circuit breakers, idempotency

Domain Knowledge
- Understanding of AI agent architectures and multi-agent systems
- Knowledge of LLM capabilities, limitations, and token economics
- Familiarity with prompt engineering and context optimization techniques
- Understanding of streaming responses and real-time data handling
- Experience with callback mechanisms and event-driven architectures
- Knowledge of data encryption and PII handling in AI contexts

Soft Skills
- Strong problem-solving ability with complex integration challenges
- Excellent written communication for documentation and tool descriptions
- Ability to design intuitive tool interfaces that LLMs can effectively use
- Collaborative mindset for working with AI engineers and product teams
- Attention to detail for schema design and error handling
- Proactive approach to monitoring and improving connector reliability
- Adaptability to rapidly evolving LLM and AI agent ecosystems

Day-to-Day Activities
- Develop new MCP connectors for enterprise system integrations
- Debug tool calling issues and optimize parameter handling for LLM consumption
- Review and improve tool descriptions for better LLM understanding
- Implement rate limiting and error handling for production robustness
- Write unit tests and integration tests for connector reliability
- Monitor connector performance and troubleshoot agent workflow failures
- Collaborate with teams on new integration requirements
- Update connectors as upstream APIs change or LLM capabilities expand

What You'll Build
- MCP servers exposing enterprise data and capabilities to AI agents
- Tool schemas and validation logic for safe LLM interactions
- Authentication and authorization layers for secure integrations
- Retry mechanisms and error recovery for resilient agent workflows
- Documentation and examples for connector usage
- Testing frameworks ensuring reliability across LLM interactions
- Monitoring and observability instrumentation for production systems

This role is ideal for engineers who are excited about the intersection of traditional software engineering and cutting-edge AI systems, and who want to build the infrastructure that enables intelligent agents to interact with real-world systems.