- Linkedin Tags: #LI-LZ
- Global Referral: No
- Division: GA Division
- Workgroup: Engineering
- Job Function: Engineer
Responsibilities
- Agent platform and orchestration
- Build and improve agent harnesses, including model and tool execution loops, context management, and structured outputs.
- Implement multi-step workflows with clear execution boundaries, task state, asynchronous dispatch, and human approval where needed.
- Enterprise systems and integrations
- Develop reusable tools and connectors for enterprise APIs, webhooks, and services, with clear input and output contracts.
- Handle data mapping, user authorization, rate limits, and integration errors; document dependencies with system owners.
- Full-stack product delivery
- Design Python backend services, secure REST APIs, and PostgreSQL data models for agent and workflow features.
- Build responsive web interfaces for task progress, user confirmations, and results using TypeScript and modern frontend frameworks.
- Security and reliability
- Protect credentials, preserve user access boundaries across systems, and apply appropriate execution isolation and audit logging.
- Improve concurrency control, retries, idempotency, cancellation, and recovery; trace failures across model calls, tools, and workers.
- Quality and collaboration
- Maintain automated and integration tests, CI/CD, and agent evaluations covering task success, output quality, latency, and cost.
Work with product, design, IT, SRE, and business system owners; contribute code reviews, documentation, and technical guidance.
Qualifications
- 5+ years of professional software development experience, including delivery of production applications across backend and frontend.
- Strong Python skills with FastAPI or similar frameworks, asynchronous programming, REST APIs, SQLAlchemy/Pydantic or equivalent tooling, and PostgreSQL data modeling.
- Hands-on experience with LLM tool calling, agent execution, or workflow orchestration; practical examples explaining design decisions, context and state management, and failure handling.
- Experience integrating enterprise or third-party systems, including API contracts, data mapping, pagination, rate limits, and retries.
- Understanding of OAuth 2.0/OIDC, authorization, role-based access control, and secure credential handling in applications.
- Ability to build reliable asynchronous workflows using task state, concurrency control, idempotency, and recovery from partial failures.
- Proficiency in TypeScript and Vue or React, including routing, state management, and reusable components. Vue 3 is preferred; willingness to adapt is essential.
- Experience with automated and integration testing, Git, CI/CD, and Docker; ability to troubleshoot using logs and metrics and evaluate agents against representative tasks.
- Clear communication with global, cross-functional teams and the ability to explain technical tradeoffs to non-technical stakeholders.
Nice to have
- Experience extending agent harnesses, SDKs, or orchestration frameworks such as LangGraph, LangChain, or custom implementations.
- Development of MCP servers, connectors, plugins, or SDKs; integrations with Microsoft Graph, Teams, SharePoint, Jira, ERP, or similar systems.
- Experience with long-running workflows, human approval, durable task state, recovery, or agent evaluation frameworks.
- Hands-on cloud deployment and operations experience, preferably AWS; familiarity with caching, queues, NoSQL databases, or vector search.
- Electron development or plugins for desktop and productivity applications.
