You choose the work, and we build and run the agents that handle it.
- You choose the work
- We build it
- You approve it
- We run it
Here’s how an agent handles a request from start to finish.
- A request comes in“Can I see this property?”
- The agent follows your rulesIt checks what’s available and offers a tour time.
- The work is recordedThe tour is booked and your team is updated.
If a request falls outside your rules, the agent asks your team and waits.
This is a leasing example. What the agent can do depends on your software and the access you approve.
You set the rules the agent follows.
What the agent does
It works only with the records you let it see, takes only the actions you allow, and keeps a log of everything it does.
What your team decides
Your team decides who gets access, what needs approval, and when to turn access off.
Which tools can it use?
Before we promise anything, we check your software plan and permissions. The agent can often work with property records, email, text messages, calendars, documents, and draft bills. Answering live phone calls is a separate project. Being able to see a balance does not mean the agent can move money.
If we can’t connect to your software directly, we look at other options, like working through a web browser with your permission, importing data, or preparing work for your team to finish. You’ll know exactly how it will work, including any manual steps, before we build.
What happens when the agent reaches a limit?
The agent can’t go past the spending limits and approval rules you set. Anything it isn’t allowed to do waits for a person. We watch for errors and keep everything working. Your current emergency coverage stays responsible for urgent problems.
The agent doesn’t go live until you’re ready.
Two callsWe learn what you need, then walk you through a proposal.
On the first call, we learn where your team needs help and tell you about Openhour. We schedule a second call, and our development team prepares a proposal. On the second call, we go over the plan, the build, and the price. If you decide to move forward, we collect the build fee. You can get it back until we have confirmed the agent can be built and you have approved that review.
BuildWe turn the way your team works into the agent’s instructions.
Once you approve that review, we build the agent around your procedures, vendors, spending limits, and approval rules. We agree on what starts each task, what counts as done, and how much of this work your team does today. The agent gets its own limited access. Anything that needs human judgment, including decisions that protect residents, stays with your team.
TestYou watch it work before it can change anything.
First we test the agent on made-up records. Then it watches your real work without changing anything. You review what worked, what didn’t, and what’s missing before you approve a small live trial on properties you choose. Your current emergency coverage stays in place.
LaunchWe run the agent, and you stay in control.
We turn the agent on after you approve, on the schedule we agreed. Openhour hosts, monitors, fixes, and improves it. AI usage costs are included in the monthly fee, which starts when the agent goes live. If a task can’t be finished, your team gets the work done so far and the reason it stopped.
Is this work worth handing off?
Work is a good fit when it comes up often, has a clear finish, and your team can set the rules for it.
When is an agent not the right fit?
An agent may not help if the work is rare, the records are hard to reach, or the decisions can’t be handed off. For simpler work, a feature already in your software may be enough. We’ll tell you if that’s the case.
Questions teams ask before we start
How is this different from using a general AI tool?
General AI tools can help with many things, but they wait for you to ask. Openhour builds and runs agents that do real work for your team, using your records and rules. They have approval limits, we monitor them, and they hand work to a person when one is needed.
Are all seven agents ready-made products?
No. They are starting designs that we build out for each company. Before we agree on a plan, we confirm your records, software access, approval rules, and what the finished task looks like. A design on this site doesn’t mean a connection to your software is already working.
What do you need from me to get started?
Tell us where your team needs help and what software you use. The first call is about your needs and getting to know Openhour, and we don’t talk pricing yet. We schedule a second call, and our development team prepares a proposal for it.
Will the agent touch my real data while you are building it?
No. We test the whole task on a made-up company first. The agent does the full job on made-up properties and residents, so any mistake it makes affects nothing real.
Do I have to change the software my team uses?
No. We aim to keep the work in the software you already use. While we plan the build, we confirm what the agent can connect to and point out anything your team would still do by hand. Your software plan and permissions can change what’s possible.
What happens when the agent cannot finish something?
Your team gets the work done so far, the reason it stopped, and what needs attention. Work that gets stuck is marked as needing help, never as finished.
Who runs the agent once it is live?
Openhour runs it. We host it, monitor it, and upgrade it to better AI models as they come out. You never have to manage a server or an AI account.
Tell us what’s taking up your team’s time.
On the first call, we’ll learn about your work and tell you about Openhour. We won’t talk about pricing yet. Then we’ll schedule a second call to walk you through a proposal.