You will own MCP (Model Context Protocol) server development for a large commerce integration program: building the governed, auditable layer that lets AI platforms and agents operate against production vendor APIs. This is frontier work with enterprise constraints: least-privilege authorization, scoped tool access, audit trails and human-approval controls are part of the design, and your patterns will need to stand up to client security review. Shipped MCP experience is a strong plus; a strong Python API engineer who has built LLM tool-calling or agent integrations and can reason rigorously about security will also succeed. This is a long-term, full-time engagement.
Key Responsibilities:
Build MCP servers that securely expose vendor APIs to AI platforms and agents
Design authorization, scoping and rate controls for agent-initiated calls against production systems
Implement logging, monitoring and guardrails so every agent action is attributable and auditable
Document MCP tools and schemas for adoption by client teams and AI platforms
Evolve the MCP security model with the lead architect as the protocol matures
Build general REST and GraphQL integrations with third-party business applications when MCP demand is light
Requirements:
5+ years of backend engineering with strong Python
MCP server development experience, or strong hands-on ability with LLM tool calling, agent tooling or AI platform integrations
REST and GraphQL API design and integration
OAuth, API authentication, webhooks, error handling and monitoring
AWS deployment with Docker and CI/CD
Security-first engineering: least privilege, audit trails, input validation, defense against prompt-driven misuse
Fluent English
Minimum 4 hours daily overlap with US Eastern time, full-time availability
Nice to Have:
Published or open-source MCP servers (share links in your application)
Experience with Anthropic, OpenAI or similar platform APIs and agent frameworks
SQL and BigQuery
Enterprise SaaS integration background
Our simple 3-step process:
Answer a few questions about your experience.
Upload your CV.
Record a brief video introduction (up to 2 minutes).
***All answers must be in English.***
FAQs:
What's the next step? If successful, you'll have one more interview before receiving a job offer if you're a good fit.
What happens after submission? We'll review your application within 2-3 business days and contact you if you qualify for the next stage.
When would I start? As soon as possible, with flexibility to accommodate your circumstances.
How long does it take? About 5-10 minutes to complete.
We look forward to reviewing your application!