About the client
The client is an international technology company developing a modular software platform for complex business operations and automated workflows. Its teams build and implement scalable solutions across different environments and use cases.
Responsibilities
•Design, build, and ship backend services and automated workflows from initial requirements through production
•Build reusable tools and integration services with clear schemas, validation, and interfaces
•Integrate third-party APIs and internal services, handling authentication, retries, rate limits, and idempotency
•Implement workflow orchestration, including state, execution limits, permissions, error handling, and human approval steps
•Build and maintain MCP servers and other tool interfaces for model-driven applications
•Develop and refine LLM instructions and tool descriptions so applications select appropriate actions and provide valid arguments
•Configure model selection, reasoning settings, tool permissions, and execution limits for each workflow
•Debug production failures, including incorrect tool selection, invalid arguments, missing actions, and incomplete execution
•Package and deploy services with Docker to cloud runtimes, with logging and instrumentation for diagnosis
•Add features and integrations to live services while protecting existing behavior through testing and regression checks
•Take initiative on architecture and reliability, identify problems early, and propose practical improvements across system boundaries
•Document services and explain their behavior to internal teams and, when needed, client engineers
Requirements
•At least 3 years of production backend software development experience
•Strong TypeScript and Node.js skills
•Practical experience integrating REST APIs and third-party systems, including authentication, structured data, and failure handling
•Docker skills and experience with at least one cloud platform: AWS, GCP, or Azure
•Hands-on experience building LLM applications that call tools, using an established SDK or framework or a custom execution loop
•Practical experience with MCP or equivalent tool-integration patterns
•Established habits around Git, automated testing, code review, and debugging production software
•Ability to make technical decisions independently, take ownership of outcomes, and communicate clearly in written English
•A degree in Computer Science or equivalent practical experience
Nice to have
•Production experience with Claude Agent SDK, the Anthropic or OpenAI APIs, LangChain/LangGraph, or Vercel AI SDK
•Experience writing MCP servers and evaluating model output with test sets, regression checks, or evaluation frameworks
•Hands-on AWS experience with ECS/Fargate, Lambda, IAM, or DynamoDB
•Experience integrating customer service platforms such as Gladly, Zendesk, Salesforce, or Amazon Connect
•Python for data work
Conditions
No conditions are stated in the posting.
