Property management AI: Four steps to doing it without the risk

Don Renyer on how to stop manual workflows from crowding out the PMC tasks that matter

One manual workflow is no big deal. The problem is that it’s never just one. 

For a mature PMC managing thousands of units, manual tasks become death by a thousand cuts. At 10,000 units, hundreds of tenants every month will experience exceptions, like unpaid rent or in-unit emergencies. 

If managers need to dig around in their PMS to resolve these issues, manual work quickly creeps out of control. Their options are limited: expand headcount, devote half their time to basic tasks, or cap their portfolio size. 

It doesn’t need to be this way. Don Renyer, Revela’s Head of Product, is here to explain how. Don leads Revela’s product strategy, including our AI rollout. To guide every step of that process, Don talks to property managers about the challenges they face at work. 

He’s figured out an airtight framework for automating these tasks so they don’t pile up—while making sure they follow business best practices. 

• Even small manual workflows can bottleneck PMC operations at scale 
• AI can help, but it must be easy to use and built around context and controls you already have
• Start by automating low-risk tasks, and recommending human action for higher-impact ones

The workflow that broke the camel’s back 

A classic example of this “manual work creep” is managing orders for maintenance or repairs. 

A tenant submits a work order. A manager then needs to read it, gauge urgency, search the vendor database for an appropriate contractor, assign the work, follow up with them, and communicate with the tenant. Again, and again, and again. 

At thousands of units, this happens countless times a day. A seemingly simple workflow becomes a full-time job in its own right.

How property management AI helps (without letting the robots run wild)

AI has the potential to nearly eliminate this problem, and it doesn’t require managers to open their books to an unknown external tool. 

To make managing thousands of homes more sustainable, AI in property management must connect to the context, workflows, and guardrails (like user permissions) you already rely on. 

It should also be simple to use for all managers. We built AI into our existing platform, so it’s accessible to anyone who knows how to use a PMS. 

“We've used Revela AI to identify 16 unassigned work orders, find the right vendors, get everything assigned, and send out the needed communications. All this took less than two minutes.” - Don Renyer

4 steps to automating maintenance work orders responsibly

AI isn’t always the answer. Start by analyzing which tasks are actually well-suited to automation, and then roll them out in a slow, thoughtful manner.

1: Rank workflow steps by downstream impact 

Identify each step in your workflow for managing work orders. Then sort each step into two categories: high-impact and low-impact. Generally, AI can assist with both types of work. The approach just needs to be a little different. 

• Low-impact: Reversible, self-contained tasks that don’t change things downstream, like tagging work orders by category or drafting messages to vendors 
• High-impact: Tasks that send a message, promise payment, or otherwise can’t be easily reversed.

“If an action doesn’t have a downstream ripple effect,  it’s typically a great candidate for full automation. Revela AI always starts by making a recommendation the user can accept or decline.” - Don Renyer

2: Automate reversible, low-risk steps 

The reversible, downstream-safe tasks you identified in step 1 can go straight to automation. To continue the work order example, you can use AI to: 

• Auto-tag work orders on intake, such as by building or type of maintenance required 
• Tag vendors with attributes, like emergency or 24-hr availability
• Enrich unit maintenance records so they're sortable and findable. 

"Every time a new work order gets submitted, AI can analyze the ticket details, enriching it with descriptive tags. There’s no risk of a destructive outcome with that.” - Don Renyer

3: Use recommendations for higher-stakes actions 

Always start with recommendations, not direct actions, for tasks with downstream impact. Revela AI was built around this framework: instead of acting autonomously, it generates a suggestion managers can approve in one click. 

• Suggest a vendor, and generate a ready-to-send email exploring their availability 
• Recommend asking the tenant for more information, and draft an email doing so

"With Revela AI, we've essentially put bumpers down on the bowling lane to make it easier for you to get a strike. You don't need AI experience to use these features, and the guardrails help ensure AI is working within boundaries you can trust.” - Don Renyer

4: Move from recommendation to action where suitable

If you’re regularly accepting AI recommendations with little to no editing, that’s a good indicator you can move to full automation. 

• AI could notify you: "Revela AI assigned work order #453 (Clogged sink) to the vendor Acme Plumbing. Vendor communications were sent and are awaiting vendor acceptance." — without you needing to execute any of these actions yourself
• Set guardrails that some actions, like deleting data or processing payment, can never be automated

"Automation should never just roll out and start making decisions for you. At Revela, we explain what an automation can do and why you might want to enable it. We think the decision to hand that work over to AI should always be intentional." - Don Renyer

PMC operations that make sense at 900 or 10,000 units 

When attention is no longer the bottleneck, headcount stops holding back growth.

Many property managers can’t hire a robust enough team to manage dozens or hundreds of these repetitive workflows a day—nor should they have to. Now, we have the technology to remove that barrier. 

If you’re looking for a PMS with AI ready to go, reach out to Revela. Our AI features were built around how property managers work.