The posting
monday has its own AI agents, built and run inside monday. Many companies are also building agents elsewhere, on Claude, OpenAI, Copilot Studio and AWS, or in n8n and Make. External Agents lets those agents work inside monday as members of the team. Each one has its own identity, permissions and activity log, and is triggered by the work happening on monday. Many companies will use both.
Connecting an agent is the easy part. Every platform offers that. The hard part is what comes after, and it's still unsolved:
- Getting enterprises to say yes. The person who wants the agent is rarely the one who can approve it.
- Getting agents to do real work. An agent that runs once is a demo. An agent that runs a process every week, with the right context and the right people around it, is a product.
- Making it a business. That means finding a model that works for customers, for sellers and for monday.
The product is live with enterprise customers. You'll own these problems, working side by side with our forward deployed engineers in customer accounts.
About The Role
Own the strategy
- Decide what External Agents becomes, what to build next and what to stop.
- Work with leadership on pricing and packaging.
Build it yourself
- Prototype agents, MCP connections and workflows before asking the team to commit.
- Build with customers and engineers, and turn what works for one customer into product for the next.
Make external agents stick
- Turn external agents that run once into agents that run a process every week.
- Make governance the reason enterprises approve them instead of blocking them.
Requirements
Core experience
- You've taken a product from 0 to 1 and can point to what changed because of you.
- You've built AI products where the model or agent is the core of the product, not a feature added on top.
- You've worked directly with enterprise customers, ideally in a forward-deployed or solutions role.
Skills that make you successful
- Building. You build working agents and prototypes yourself, using tools like Claude Code, Cursor or n8n.
- Agent integration. You know MCP, APIs, webhooks, OAuth and permissions well.
- Building with customers. You can sit with a customer and build a working solution in the room.
Nice to have
- Founder or early-stage startup experience.
- Background in workflow automation, identity or enterprise security.
Why this role
- Open problems. How outside agents should work inside a company's systems is still being figured out, and you'll be the one deciding.
- Real customers. This isn't a lab project. Enterprise customers already use it, and what you ship lands in their workflows right away.
- Hands-on. You'll build, not manage a roadmap, with the scope of a startup and monday's enterprise customers behind you.
What monday.com http://monday.com can offer you:
- Join a customer-obsessed team where transparency and ownership are how we actually operate - not just things we put on a wall
- Work on a multi-product suite powered by advanced AI, where what you ship reaches hundreds of thousands of customers worldwide
- Competitive salary, benefits, bonus potential, and equity eligibility for many roles — we believe people should share in our success
- Grow fast through a dedicated L&D team, AI workshops, bootcamps, and internal hackathons - the industry moves quick and so do we
- A collaborative, global culture backed by Employee Resource Groups and workplace teams across New York, Tel Aviv, London, Munich, Warsaw, Tokyo, Sydney, and more



