高级软件工程师，内部部署产品
Senior Software Engineer, Internally Deployed Products
在 ClickUp，我们正在打造未来的工作方式：首个真正融合 AI 的工作空间，将任务、文档、聊天、日历和企业搜索统一起来，所有功能都由以情境为中心的 AI 驱动。我们是一家以 AI 为核心的企业。每位团队成员都需要每天使用 AI，并且我们在招聘过程中会评估 AI 的熟练程度。加入我们，一起重新定义可能的边界。🚀
使命
Foundry 是 ClickUp 内部的 AI 创新实验室——嵌入在 GTM 系统中，负责将 AI 能力转化为可生产部署的内部产品，使每个 GTM 功能更快、更智能。我们构建支撑 Sales、Marketing、Post-Sales 和 Revenue Operations 中 AI 优先工作的基础设施。
作为该团队的高级软件工程师，你将负责 MCP 服务器平台、代理编排层和内部工具的工程技术交付——发布每天被数百名 ClickUp 员工使用的生产系统，并通过将 AI 工具视为第一工程合作者来提升自己的处理能力。
你将负责的内容
MCP 服务器平台
- 设计、构建和运营 Model Context Protocol 服务器，向 GTM 堆栈中的 AI 代理暴露 CRM、工单、分析和通信数据
- 实现 Okta PKCE 认证流程和基于角色的访问控制（RBAC）策略执行，确保代理只能访问其授权的数据
- 维护 AWS 上的部署基础设施（Bedrock、Lambda、ECS、API Gateway），并在适用时为 GCP 工作负载做出贡献
- 负责可观测性：结构化日志、分布式追踪、延迟 SLO 和每个生产服务器的值班手册
代理编排与 AI 原生产品
- 构建和维护多步骤自主代理，执行端到端的 GTM 工作流——线索资格审核、交易室组装、入职自动化、支持分类等
- 架构提示工程框架、工具调用模式和代理评估工具，使 AI 行为可预测且可审计
- 与 LLM 提供商（Anthropic、OpenAI、AWS Bedrock AgentCore）集成，并维护版本锁定、成本跟踪的模型配置
- 交付 AI 驱动的内部应用程序（网页应用、CLI 工具、Slack 集成），让非技术型 GTM 利益相关者无摩擦地使用
GTM 平台工程
- 负责 TypeScript/Node.js 后端和 React/TypeScript 前端的全栈功能交付，用于内部工具
- 编写 Python 自动化脚本，
查看英文原文
At ClickUp, we're building the future of work: the first truly converged AI workspace unifying tasks, docs, chat, calendar, and enterprise search, all supercharged by context-driven AI. We are an AI-native company. Every team member is expected to leverage AI daily, and we evaluate AI fluency as part of our hiring process. Join us and help redefine what's possible. 🚀
THE MISSION
The Foundry is ClickUp's internal AI innovation lab — embedded inside GTM Systems and accountable for turning AI capabilities into production-grade, internally deployed products that make every GTM function faster and smarter. We build the infrastructure that powers AI-first work across Sales, Marketing, Post-Sales, and Revenue Operations.
As the Senior Software Engineer on this team you will own the technical delivery of our MCP server platform, agent orchestration layer, and internal tooling — shipping production systems used daily by hundreds of ClickUp employees, and scaling your own throughput by treating AI tools as first-class engineering collaborators.
WHAT YOU'LL OWN
MCP SERVER PLATFORM
- Design, build, and operate Model Context Protocol servers that expose CRM, ticketing, analytics, and communication data to AI agents across the GTM stack
- Implement Okta PKCE authentication flows and RBAC policy enforcement so agents access only the data they're authorized to touch
- Maintain deployment infrastructure on AWS (Bedrock, Lambda, ECS, API Gateway) and contribute to GCP workloads where applicable
- Own observability: structured logging, distributed tracing, latency SLOs, and on-call runbooks for every production server
AGENT ORCHESTRATION & AI-NATIVE PRODUCTS
- Build and maintain multi-step autonomous agents that execute end-to-end GTM workflows — lead qualification, deal room assembly, onboarding automation, support triage, and more
- Architect prompt engineering frameworks, tool-call schemas, and agent evaluation harnesses that make AI behavior predictable and auditable
- Integrate with LLM providers (Anthropic, OpenAI, AWS Bedrock AgentCore) and maintain version-pinned, cost-tracked model configurations
- Deliver AI-powered internal applications (web apps, CLI tools, Slack integrations) that non-technical GTM stakeholders use without friction
GTM PLATFORM ENGINEERING
- Own full-stack feature delivery across TypeScript/Node.js backends and React/TypeScript frontends for internal tooling
- Write Python automation scripts, ETL pipelines, and data transformation layers that feed GTM analytics and AI context
- Collaborate with Systems Engineering and GTM Engineering teams on cross-cutting API standards, data contracts, and integration patterns
- Conduct code reviews, establish engineering standards, and actively mentor junior engineers toward higher leverage
AI-NATIVE DEVELOPMENT PRACTICE
- Use AI coding assistants (Claude, Cursor, GitHub Copilot) as primary engineering accelerators — not supplements — to ship at a pace that punches above a single engineer's weight
- Document AI usage patterns, prompt templates, and agentic workflows so the team's collective throughput compounds
- Stay current on MCP protocol evolution, agent frameworks (LangGraph, CrewAI, custom), and emerging LLM capabilities; bring back what matters
REQUIRED QUALIFICATIONS
- 5+ years of professional software engineering experience with production systems
- Expert-level TypeScript and Node.js — idiomatic, typed, testable server-side code
- Strong Python — automation scripts, data pipelines, and scripting for AI/ML tooling
- Meaningful AWS deployment experience: Lambda, Bedrock, ECS/Fargate, API Gateway, IAM, Secrets Manager, CloudWatch
- Demonstrated experience integrating with LLM APIs (OpenAI, Anthropic, AWS Bedrock, or equivalent) and shipping AI-powered features to real users
- Solid foundation in REST API design, OAuth 2.0 / OIDC authentication, and secure credential management
- Experience with CI/CD pipelines, infrastructure-as-code (Terraform, CDK, or SAM), and cloud cost awareness
- Clear written communication: design docs, ADRs, and runbooks that others actually read
- Track record of using AI tools (LLM assistants, copilots, agentic workflows) as a genuine productivity multiplier — not a gimmick
STRONGLY PREFERRED
- Hands-on experience with MCP (Model Context Protocol) — building servers, defining tool schemas, or operating multi-server agent environments
- GCP experience (Cloud Run, BigQuery, Cloud Functions) to complement AWS work
- Salesforce, HubSpot, or other CRM API integration work — understanding GTM data models is a significant advantage
- React or Next.js front-end capability to ship internal dashboards and tooling without hand-offs
- Familiarity with agent orchestration frameworks: LangGraph, AutoGen, CrewAI, or custom orchestration patterns
- Experience in a GTM Systems, RevOps, or Sales Engineering context
- Okta / identity provider integration work (PKCE flows, SCIM, token management)
HOW WE OPERATE
The Foundry runs on three principles that every team member is expected to embody:
- Ownership — You own the full lifecycle of what you build: design, delivery, observability, and iteration. We don't have a separate ops hand-off.
- Velocity — We ship to production in days, not quarters. AI tools are how we maintain that pace without sacrificing quality.
- Atomicity — Every deliverable is a discrete, composable unit — an MCP server, an agent, an API. Nothing we build requires everything else to work.
WHO THRIVES HERE
This role is a strong fit if you are genuinely excited to work at the intersection of AI infrastructure and internal product — not just as a user of AI, but as someone building the layer that makes AI useful for an entire go-to-market organization. You are comfortable with ambiguity, you default to shipping, and you see every piece of internal tooling as a product that deserves operational discipline.
This role is not a fit if you need heavily defined requirements before starting work, prefer long planning cycles to iterative delivery, or view AI tools as optional extras rather than core to how you operate.
Equal Opportunity Employer
ClickUp is an Equal Opportunity Employer, and qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, or national origin.
Privacy Notice
ClickUp collects and processes personal data in accordance with applicable data protection laws. You can find further details by viewing our Global Candidate Privacy Notice. https://clickup.com/careers-terms/candidate-privacy-notice
If you are a Philippine Job Applicant, please also see our Philippine Data Privacy Notice https://t333.s.clickup-attachments-stg.com/t333/d9a53ce2-6d2c-48cd-84b1-ed4e525f9613/Philippine%20Data%20Privacy%20Notice_2024.pdf?view=open for further details.
Visa Sponsorship
Please note we are unable to sponsor or take over sponsorship of an employment visa for roles outside of engineering and product at this time. Sponsorship for engineering and product roles is not guaranteed, but is instead based on the business needs for that specific role at that time. Please reach out to the recruiter with any questions.
Fraud Alert
ClickUp Talent Acquisition will only initiate contact via an @clickup.com http://clickup.com email or through our official careers portal on clickup.com http://clickup.com. We will never request fees, payments, or sensitive personal information. Please disregard any offers received outside these channels and report them to support@clickup.com.
AI Processing Notice
ClickUp may use artificial intelligence and machine learning technologies to help review and screen candidates' employment applications against role-related criteria. These tools support, but do not replace, human decision‑making. If you have questions or need an accommodation in the recruitment process, please contact us at AskPeople@ClickUp.com.