“AI, for us, is about improving resident experience,” says Philip Irby, CTO at AMH. The company (formerly American Homes for Rent) introduced an AI-powered virtual assistant named Amy that has transformed customer relations in the call center.
Amy can handle basic inquiries around things like availability, while simultaneously untethering employees from endless routine calls, so they can instead focus on more complex inquiries and tasks. She answers the phone on the first ring — every time — engaging in a two-way conversation and following it up via text and email.
Box COO Jon Herstein talked to Irby for the Box AI-First podcast. This episode, recorded live at BoxWorks 2025, sheds light on how AMH uses AI, cloud technology in general, and Box integrations in particular to streamline property management operations and elevate the resident experience.
Key takeaways:
- Centralize unstructured content to enable effective AI agents that improve resident experience and operational efficiency
- Pair AI deployment with disciplined change management and long-term ROI thinking so employees adopt the systems and AI can free staff for higher‑value work
- Build targeted, agentic AI integrations so automation can handle routine queries and trigger human workflows when needed
Increasing both customer and employee happiness
Managing 62,000 single-family homes creates a lot of customer service calls every day. AMH leverages AI to ensure that both current residents and people looking to rent get their questions answered quickly. Irby says customers love Amy, and frequently leave positive reviews about their interactions with her.
Amy also enhances operational efficiency by freeing employees up to handle more complex tasks. As Irby says, “They aren't answering questions that could be easily answered, like whether a house is available. Amy knows the answers to those questions."
As a result, that employee now has more time to attend to more complex customer issues. In the past, an escalated call might mean a 15-minute wait to talk to a manager. Now, that manager’s time has been greatly freed up, so they are available sooner. It’s a virtuous customer service and employee satisfaction cycle.
AMH has big plans for Amy, like using agentic AI to connect her to specific agents within the company. For instance, Amy could connect to the lease agent to answer granular and individual lease questions, or the HOA agent to answer “When is my trash day?” or “Can I park on the street on Tuesdays?” If the issue requires human assistance, like a maintenance request, the agent could set that particular workflow in motion.
The end-of-the-day goal: quick assistance for residents, simultaneously relieving human workers from the burden of having to look through paperwork and files for the info.
Centralizing content for AI readiness
Herstein posed the question, "If your content is still scattered everywhere, how are you going to have a unified AI strategy?"
One of the foundational strategies AMH employs to maximize value from AI is centralized content. As Irby says,
We're all good at structured data. Right? We're terrible at unstructured data. Getting everything into one place is the first step.
Centralized content is a universal challenge for almost every company of any size. AMH is only about 13 years old, so not a traditional “legacy” company, but because it has grown with an entrepreneurial mindset — at times buying thousands of houses a month — there has innately been a certain amount of content chaos. Irby describes, “There were a lot of disconnected processes, lots of emails-as-workflow, that kind of stuff. The legacy we had was all those disparate workflows.”
Irby was instrumental in building a custom platform specific to the business case of single-family home rentals. This included installing Box for all the underlying unstructured data. Today, AMH doesn’t own a single piece of hardware in a data center anywhere; it’s all in the cloud. He has also instituted a deliberate migration away from legacy systems into Box integrations:
We've retired SharePoint. We've retired all Google Drive stuff. Everything we have is in Box. Having done that, we stand on the precipice of getting all the value we can out of our documents.
This content-first approach enables AMH to consolidate unstructured data — historically a challenging feat in any industry — and unlock its potential for AI learning and decision-making.
Custom building on Box with open APIs
Open APIs are essential to how AMH has built its content stack. “There are certain things we don’t ever want to do,” explains Irby, “like build a content management system. What Box does is always going to be so far ahead of us that it takes a huge burden off of us to do foundational stuff.”
So rather than build a bespoke content management system, his team built a microservice that interfaces with Box to componentize document management, using Box UI components to deliver micro front-ends in their resident portal. Then, Irby says, “We built the parts that are unique to us,” like tools specific to the single-family rental business.
In general, any time Irby’s team evaluates a new technology vendor, the first filter is the question “Do you offer open APIs?”
Giving agentic AI the best possible context
For AI to be useful, it needs context. And a lot of context naturally happens in email.
Irby has a strategic goal of getting all existing and ongoing company email conversations into a Box Hub in order to use that incredibly valuable collection of content to further train and inform AI agents. In addition, by consolidating entire threads with each resident, and applying AI to that trove of content, it will become much easier to have continuous conversations with residents over time and solve their problems more quickly.
Another transformative use case is AMH’s contract management process. The company has a home-building arm that deals with thousands of contracts annually, all coming in through Box. The CSO currently has to go through each contract individually, but Irby’s team is in the process of building AI agents to automatically review these documents for compliance and efficiency as part of the standard workflow. Irby says, "You could train AI to know what we need to look for — and then it just does it."
These efforts showcase the versatility of AI in tackling complex, time-consuming tasks across multiple domains.
Change management as a cornerstone of AI innovation
AI implementation has to come from business goals. For AMH, a singular goal was to create a stellar resident experience, and that’s how they’ve made decisions on where to use AI. Irby says, “There’s so much technology and so many rabbit holes for people to go down. You have to have a clear North Star and be disciplined to not get distracted.”
He also emphasizes how important it is to transition out of legacy systems “as fast as you can” and “bring everyone with you.” Right away, he insisted that all groups within the company put all their content on Box. “That wasn’t always easy,” he says, “and I had to have some hard conversations. You have to figure out ways to engage not only your IT organization but your employees so they find value in it. Once you get that grassroots stuff going, then it’s word of mouth.”
His advice underscores the massive cultural and structural changes AI demands, highlighting the need for inclusive change management: "AI is not just the biggest technological thing that's happening right now. It's also the biggest change management... the pace of it is now different than anything that's gone before."
Investing in ROI for the long game
While using AI to fulfill business goals might be top priority, it’s not always a straight path to ROI. “Sometimes,” Irby says, “it’s an investment. Sometimes we spend more money to make the resident experience better, and we don’t expect to get it back in the short term.” Instead, they might measure whether the metric “return residents” goes up over time, or another type of metric that shows the long-term ROI of decisions.
Ultimately, per Irby: “It’s not just about doing the same things you were doing more efficiently. It’s about doing things you could never do before.” By focusing on innovative tools, refining change management processes, and investing in centralized systems, companies can chart a sustainable and impactful AI journey.
Catch the full episode
Ready to dive deeper into this discussion? Don’t miss episode 6 of the AI-First Podcast: How AI powers every touchpoint at AMH. Subscribe now to stay informed and get inspired about the AI-first era.
