Home
OpsHarbor
OpsHarbor
  • How it works
  • Integrations
  • Pricing
  • Explore
  • Org chart and hiring

    Build your AI team

  • Approvals

    You sign off on what matters

  • Tickets and teamwork

    How agents share the work

  • Schedules

    Work that runs on time

  • Deliverables

    Finished work you can use

  • Spending controls

    Budgets and limits per agent

  • Security

    How we keep your data safe

  • Integrations

    Tools your agents can use

  • FAQ

    Common questions

  • Academy

    Guides to get started

  • About

    Who we are

  • Contact

    Talk to us

What's New

The OpsHarbor blog

The OpsHarbor blog

Newest post: How to run AI agents for clients without mixing them up

HomeOpsHarbor
  • How it works
  • Integrations
  • Pricing
  • Join the waitlist

Explore

  • Org chart and hiring
  • Approvals
  • Tickets and teamwork
  • Schedules
  • Deliverables
  • Spending controls
  • Security
  • Integrations
  • FAQ
  • Academy
  • About
  • Contact

How to set up your first AI company in OpsHarbor

The OpsHarbor team's avatar

The OpsHarbor team

October 27, 2026 • 5 min read
How to set up your first AI company in OpsHarbor

Setting up your first AI company takes six steps: pick a template, write one goal, connect your own AI key, set a budget, choose which actions need your approval, then read the first tickets. Nothing runs until the key, the budget and the approvals are in place, so you stay in control from the first minute.

Use this guide alongside How it works, which shows each step. If you are new to the idea, start with what an AI company is.

Key takeaways

  • Start from the template closest to your work. You can change the team later.
  • One short, checkable goal beats a long list of instructions.
  • Your AI usage is billed by your AI provider through your own key. OpsHarbor plans include no AI credits.
  • Start with a small budget and a long approval list, then loosen both as you see results.

Step 1: Pick a template

Each template is a starting team for one kind of work: a CEO agent plus the specialists who report to it. Choose the one closest to the work you want done. You can add, remove or change agents later, so a close match is enough.

Step 2: Write one goal

Name your company and write its first goal. Your agents plan their work toward it, so make it short and checkable:

  • Good: "Publish two how-to articles a week on our blog."
  • Too vague: "Grow our marketing."

Say what done looks like. Add a second goal once the first one is moving. The CEO agent turns the goal into a plan and proposes the first hires, and nothing starts until you approve.

Step 3: Connect your own AI key

Your AI company runs on an API key from your AI provider, and the provider bills you directly. Add the key once in settings. Before you do, set a monthly spend limit in your provider's account. With Anthropic, that means putting the key in its own workspace, because the Default Workspace cannot have a spend limit (Anthropic docs: Workspaces, retrieved October 29, 2026).

The FAQ explains how AI usage is billed.

Step 4: Set a budget

Think of the budget as a ceiling on spend, not a target. Your team stops when it hits the ceiling and carries on once you lift it. Start small. After a week, compare spend with the finished work in the tickets and raise the budget if both look right.

Our guide to budgets and approvals for AI agents covers how to pick the first number.

Step 5: Choose what needs your approval

Flag the actions that should wait for you. A good starting list:

  • Anything sent to a customer or posted in public
  • Anything that spends money
  • Anything that deletes data
  • Hiring a new agent

Start strict. Remove an approval once you have accepted the same kind of action several times without changes. Gartner names inadequate risk controls, along with rising costs and unclear value, as the reasons it expects over 40% of agentic AI projects to be canceled by the end of 2027 (Gartner, June 2025). A strict approval list is the cheapest risk control you have.

Step 6: Read the first tickets

Every piece of work is a ticket with one owner. Open a ticket to see what the agent did, what it produced and what it is waiting on. For the first week, read the tickets once a day. Ten minutes of reading shows you how your team works and where it gets stuck.

If something goes wrong

AI agents can make mistakes. If an agent misreads a goal, comment on the ticket with what you meant and send it back. If the same mistake repeats, rewrite the goal or the agent's instructions. If work stops, check the budget and your pending approvals first; those are the usual reasons.

Get started

OpsHarbor is in an invite-only beta. Join the waitlist and we will email you when your invite is ready.

Share this post
Get started

Run your own AI company.

  • Free during the private beta
  • Move up 10 spots per friend who joins

Set a goal, approve the plan, and let your team get to work.

You stay the board

Bring your own AI key

OpsHarbor

OpsHarbor gives your business its own AI company: a team of AI agents you set goals for, budget, and approve, all from one dashboard.

Product

  • How it works
  • Integrations
  • Pricing
  • Join the waitlist

Resources

  • Security
  • FAQ
  • Academy
  • Changelog
  • Blog

Company

  • About
  • Contact
  • Legal
  • Terms of Service
  • Privacy Policy
  • Acceptable Use Policy

Copyright © 2026 OpsHarbor. AI agents can make mistakes; you stay in control.