Register for NexusCon 2025!
Technology for facility maintenance has long struggled with engagement from the boots on the ground. Platforms like computerized maintenance management systems (CMMS) promise a “single source of truth” for work orders and equipment data, but too often, the people doing the actual work barely use them.
That’s a huge problem for building owners spending money on these tools—if technicians aren’t inputting data, the system can’t reflect reality. The bigger problem is what this gap represents: ineffective maintenance programs leading to missed work, undocumented issues, and reactive repairs.
One Accruent survey found that only 39% of businesses consistently use a CMMS to track maintenance tasks (the rest resort to spreadsheets or even pen and paper). This low adoption means many preventive tasks fall through the cracks, contributing to growing maintenance backlogs and costly breakdowns.
There are some good reasons for this poor engagement:
In short, today’s tools have not been very user-centric for the people on the ground, so those people ignore them—and the hoped-for efficiency and data insights never materialize.
Now that’s starting to change. Gamification, AI, and behavioral psychology principles influence users to return frequently, spend more time on the platform, and even enjoy doing so. Vendors aim to make facility software “sticky”—something folks habitually use, not dutifully avoid.
If that playbook sounds familiar, it’s because it’s the same engagement strategy that consumer tech (especially social media) has mastered over the last 20 years. Social media platforms leverage gamification tactics like:
These techniques, as described by author Nir Eyal’s popular Hook Model, create habit-forming loops of trigger → action → variable reward → investment. In other words, they cue you to engage, reward you in unpredictable ways that keep you guessing, and encourage you to put something of yourself into the system (time, effort, personal data) so you’re more invested in coming back.
While we can debate the merits of endlessly scrolling Instagram or chasing likes, there’s a growing case that some of these tactics can be repurposed to solve the long-standing engagement gap in building maintenance tech. What made Facebook addictive could make a maintenance app finally stick—helping buildings be better maintained, saving money, and improving occupant satisfaction (which can drive higher revenues for building owners through retention and reputation).
What’s especially exciting is the addition of modern AI into this formula. Natural language chat interfaces and AI assistants are supercharging the engagement strategy—and the capabilities of technicians and operators, too. A chat-based AI can draw users in with a conversational, almost human interaction, and then help them get things done faster than ever.
And this boost comes at a critical time: facility teams are being asked to do more with less, right as a generation of experienced technicians retires and a skilled labor shortage hits the industry. In the U.S., as Baby Boomers retire, 62% of companies are already struggling to fill skilled trade positions left behind. Maintenance departments everywhere feel understaffed and overextended. In this context, tools that actually engage the workforce—and extend their abilities with AI—aren’t just nice to have; they might be essential to keep buildings running reliably.
Facing these challenges, smart building technology vendors are taking pages from the gaming and social media handbook to better engage every persona that interacts with their systems. The key is recognizing that different users have different motivations.
As Jonathan Kroll, Co-founder and Chief Product Officer of Visitt, told us, “It’s gotta start with the user’s motivations, which are different for different personas.” A property manager, for example, already has a built-in incentive to address maintenance issues—it helps keep their tenants happy and makes them look good. But technicians and field operators? That’s a different story.
Kroll points out that technicians’ days are hectic and dynamic; stopping to fill out a digital form after fixing a toilet or AC unit can feel like just another task with no immediate benefit. But since this maintenance data is critical, the question becomes: how do you meet the technician where they’re at and give them a reason to use the system more?
One tactic is to make the software personally rewarding to use, beyond just the abstract promise that “data entry helps your company.” For example, Kroll mentioned the concept of an “inbox zero” achievement for work orders. This plays on a simple psychological urge—the satisfaction of completing a list and seeing it cleared. “We all want to clean our to-do list,” Kroll said.
When a technician resolves all their assigned work orders (or even just all the high-priority ones) for the day, the app might celebrate that with a big green checkmark, some on-screen confetti, or a congratulatory message. It’s a small dopamine hit, but it feels good—a moment of accomplishment instead of drudgery.
Another strategy is recognition and visibility. Technicians often express frustration that they “do a ton of work but nobody sees it.” In a busy facility, if an HVAC tech prevents a disaster or works through the night, tenants and bosses might never know—until something goes wrong. To address this, Visitt allows the end customer (e.g. the tenant) to provide quick feedback or ratings on a completed request. Kroll described how Visitt lets tenants give a simple 1–5 star review or a thumbs-up when their issue is resolved. That feedback goes straight to the technician’s profile. It’s also “celebrated” in the app, meaning the tech (and their team) might see a star count or a note like “Tenant X praised your response time!”
Over time, accumulating a high satisfaction rating becomes a point of pride. Management can even set a KPI like “maintain at least a 90% positive tenant rating” for the team, which gamifies the customer service aspect of maintenance. The important nuance: this isn’t about punitive metrics or public shaming; it’s about rewarding the frontline with recognition that historically they lacked.
Visually, the interfaces are also borrowing from consumer app design to make the experience more engaging. Instead of dry tables and forms, you’ll see friendly graphics when a task is completed or a progress gauge of how close you are to some target. These little UI touches act like the “likes” and badges of a maintenance app—small rewards or status signals that make the software less sterile.
Visitt created a “score” for building performance that updates based on maintenance activities and outstanding tasks. “Once you introduce the score, they want to get 100,” Kroll observed, noting how property managers and even executives got hooked on checking and improving their building’s score. It’s essentially a leadership board for buildings: if your property is sitting at a B grade (say 80/100) in maintenance health (maybe based on open issues, response times, etc.), that creates a natural drive to push it to an A.
Gamification isn’t just for the in-house staff—it’s also being used to engage service vendors and contractors who work on buildings. Bob French of 75F gave an example of how they’re applying game design to the commissioning and maintenance of HVAC control systems. 75F recently launched a feature called “Easy Street” which, as French explains, “gives a site installation score and gamifies the commissioning process.”
Here’s how it works: when a contractor installs 75F’s HVAC controls in a building, they run the Easy Street automated checkout. The system then analyzes whether all controllers and sensors are installed correctly and calibrated. Instead of just spitting out a dry report of issues, it produces a score—essentially rating the quality of the install. If you, the installer, got an 86 out of 100, you know there’s room for improvement. The tool also generates a punch list of what needs fixing (maybe a sensor offline or a config not optimized).
They can address the punch list items and run it again to get closer to that coveted 100. 75F envisions technicians really taking to this, even to the point that “if someone gets a good score then they will go on LinkedIn and brag about it.” In other words, turning a perfect installation into a badge of honor in the professional community. The score gives immediate feedback and a sense of achievement for doing a thorough job, which is much more motivating than a long checklist with no clear indicator of success.
That same scoring concept continues after installation as well. A facility manager can periodically run the Easy Street diagnostic on their building (whether monthly or quarterly) to see if performance is drifting. They’ll get a score each time, and perhaps more importantly, see the trend. Maybe last quarter the building was a 92, and now it’s down to 88—time to investigate and bring it back up.
This creates an ongoing game of keeping the building performing at its best. The historical log of scores taps into our natural desire to improve our “personal best” and not let it drop. French expects facility teams to take pride in maintaining high scores, which in practice means proactively fixing issues the software identifies.
Jean-Simon Venne of BrainBox AI agrees: make operating a building feel like playing a game. Instead of discrete rewards like badges or points, he says “it’s about performance… Can I get closer to 100%?” Venne described how any building system (say a chiller plant) has an optimal performance curve—essentially the most efficient it could be running under current conditions. If you give the operator a real-time metric of how close they are to the optimal, it becomes a challenge to push it closer.
Crucially, Venne also highlighted why current tools struggle to engage: chaotic UI and information overload. Traditional building automation systems and cloud applications often bombard users with data—dozens of tabs, graphs, and raw values for every piece of equipment. “The problem… is that traditional software UIs have so many tabs. The more data you have, the more tabs you have,” he noted.
A veteran operator might navigate that, but your average maintenance tech or property manager just finds it bewildering. “You just want to know, is my building okay or not? Should I send a truck or not?” Users need simple answers to important questions, not a sea of charts.
Gamifying operations means streamlining the experience so it’s easy to consume (games usually teach you gradually and keep the interface intuitive). If a tool can tell an operator “All systems go—you’re at 95% optimal” or “Alert: you’re dropping—here’s what to do”, that concise insight keeps them engaged far better than dumping a PDF report on their lap.
The next evolution of this trend is weaving AI and natural language interfaces into building operations, turning the whole experience even more interactive and personalized. If gamification provides the motivation to use the tool, AI is providing a new means to use it—one that can drastically reduce friction and boost productivity.
An AI assistant changes the game (pun intended) by making the software feel less like software at all. Instead of rigid menus and forms, users get a conversation partner that can understand requests and guide them to solutions. This adds a new layer to the habit-forming loop: the system isn’t just reactive, it actively engages with you, almost like another teammate.
Think of how people started engaging with ChatGPT and similar AI assistants in the past couple years. Millions of users found themselves coming back again and again to these bots to ask another question, try another idea—partly because it’s so easy (just type what you want, no training needed), and partly because the responses are varied and sometimes surprisingly helpful. It’s the Hook Model in action: you have a question (trigger), you prompt the AI (action), you get an answer that’s sometimes great, sometimes mediocre (variable reward), and then you refine your approach or give feedback (investment), which makes the next interaction a bit better.
It can become addictive in a useful way. In fact, ChatGPT became the fastest-growing app in history, reaching 100 million users in just 2 months—a testament to how compelling an AI-driven interface can be when it works. Now apply that concept to a building management context: instead of digging through manuals or ignoring complex software, what if a building operator could just ask an AI assistant “Hey, why is it so hot on the 3rd floor today?” and get a credible answer? That’s where we’re headed, and it builds on all the gamification groundwork described earlier.
Log in or join Nexus Pro to read on... we'll cover exactly how AI chat interfaces work to enable this and what it means for FM.
The AI makes the action part of the habit loop incredibly simple (“just ask me”), and it can provide a tailored reward (the exact info or fix needed) much faster than a human could alone. Moreover, it can make the interaction feel engaging—even fun—because it’s a bit like talking to a super-knowledgeable colleague or playing around with a friendly oracle.
Jean-Simon Venne calls this approach “AI for UI,” meaning using artificial intelligence to reinvent the user interface. BrainBox AI’s interface, nicknamed “ARIA,” is designed so that “you’re basically going to talk to your screen rather than [use] mouse and keyboard,” Venne said. He likened it to JARVIS from the Iron Man films—essentially an AI butler for the facility.
Thanks to advances in voice recognition and large language models, the barriers that made this idea sci-fi a decade ago are largely gone. Venne noted that issues like AI hallucinations and slow response times, which were concerns initially, are being overcome with new techniques (for instance, grounding the AI in the building’s own data to keep it factual).
BrainBox AI had a head start here: “we built an agent with AWS before [agent] was a common term,” he said, describing how they trained their AI on a normalized dataset of building telemetry. They’ve since expanded it by ingesting equipment PDFs (manuals, specs) and comparing live sensor data to known performance curves. The result is an assistant that can answer complex operational questions by drawing on everything from design intent to maintenance history.
Venne gave an illustrative scenario of what’s now a beta feature: A large customer site in Louisiana calls in saying, “It’s too hot in here.” An operator getting that complaint might normally have to log into a building automation system (BAS), sift through trends for that zone, check the HVAC unit, maybe remote into a controls interface—a process that could take a couple of hours to pinpoint the issue (if it’s even found). In the AI-assisted world, that operator instead tells the AI, “Building 4, Zone 3 is too hot. What’s going on?”
The AI instantly parses this (understands the question and context—perhaps it knows the weather, the recent sensor readings, and that a work order was done on that unit last week). It then investigates across the data: it might pull the HVAC unit’s latest telemetry, cross-reference the maintenance log (e.g. a filter was changed last week), recall that the manufacturer manual suggests a certain error code for the symptoms, and even note that “oh, the optimal setpoint for this condition should be X and it’s currently set to Y.”
Within a couple of minutes, it returns an answer: “The cooling valve on AHU-7 appears to be stuck closed, causing the temperature rise in Zone 3. I recommend checking that valve; a temporary override has been applied to reduce supply air temperature.” In essence, it did the 2-hour investigation in 5 minutes and handed the tech a clear next step. As Venne put it, “You can do a lot more of these investigations than you can do today.” The technician’s capacity is supercharged—they can respond to many more complaints per day because the heavy diagnostic work is automated.
This intelligent assistant approach doesn’t just speed up individual tasks; it helps bridge the skills gap on the team. A junior tech with only a year on the job can now handle an issue that previously might have required a 20-year veteran’s intuition. The AI, having been “trained with a lot of HVAC documents” and past data, can infer the cause with “the same logic that a human would have followed,” according to Venne.
In other words, AI can act as a force multiplier for a shorthanded maintenance team. It won’t replace the humans (you still need someone to turn the wrench on that stuck valve or to verify the AI is correct), but it can guide less experienced staff to do more on their own and help senior team members accomplish more faster. Even things like suggesting replacement parts or optimal settings can be handled, reducing the cognitive load on the human and letting them focus on execution and higher value tasks.
75F’s Bob French shared a similar vision from the controls standpoint, which they’ve begun to realize in their product called “Saffron” (a nod to a helpful assistant). “You can talk to your building now,” French said. The intended user for Saffron isn’t the typical controls engineer—it’s “not a technical persona.” In fact, they talk about the concept of “the zero interface,” where an end user might not even know a building automation system is involved.
Imagine an office manager or a tenant who has access to a simple chat window or voice assistant for their building. If something’s wrong, they just state it in plain English: “The meeting room is too cold,” or “The lights on our floor are flickering.” Saffron takes that input, interprets it, and interfaces with the building’s systems behind the scenes. The user doesn’t need to learn any dashboard or terminology.
French gave a concrete example: “We’re calling for cooling, and the air coming out isn’t cool.” A non-technical person might report that observation. Traditionally, that would trigger a work order and likely a truck roll—a technician driving out to the site to see why the AC isn’t cooling. Often, the fix might be something as trivial as a setpoint that was set incorrectly or a system that needed a reset—tasks that don’t actually require a wrench, just knowledge of the control system.
With an AI assistant like Saffron, the system itself can recognize this scenario. It sees that the zone temperature is high despite a cooling command; it checks the equipment status and might discover that, say, the cooling setpoint was raised or a damper is closed. It can then either automatically correct the issue or guide the user: “I’ve adjusted the setting; give it a few minutes and it should start cooling. Let me know if it’s not better.” This avoids an unnecessary service call.
The maintenance team didn’t have to drop everything to drive over for a 30-second fix; they can reserve their time for truly complex issues or tasks that require hands-on work. This illustrates how AI can empower even tenants or non-engineers to solve minor building issues, which is another form of engagement—it expands who can interact with building systems beyond the typical facilities team.
Importantly, what 75F and others are doing here is reducing the training and expertise barrier to using building technology. “Prior to Saffron, [if] they needed end user training to implement a BAS… now, you can just give those commands to the AI and it does it for you,” French explained.
Historically, one reason smaller buildings (and their staff) never utilized advanced energy management or automation systems is that those systems were too complex. No one had the time (or inclination) to learn the user interface or interpret the data, so they’d give up. By hiding all that complexity behind a conversational AI, these vendors are essentially offering a “fluent translator” between humans and building machines. You don’t need to know which submenu has the air handler controls; you just say what outcome you want.
This means building technology can actually be used in the many places it wasn’t before. It’s a bit like how early computers were only used by specialists, until graphical interfaces and voice assistants made them accessible to everyone.
For years, the smart buildings industry has struggled with a paradox: there’s more and more tech available to run facilities better, but the people on-site often don’t use it.
Technicians, property managers, tenants… they’re all people, and people respond to feedback, challenge, and easy-to-use helpers. Giving a maintenance tech a pat on the back for a job well done (even if it’s just a digital high-five), or a challenge to beat their past performance, can turn a mundane workflow into something engaging.
Crucially, this convergence of gamification and AI could bring smart building benefits to the “long tail” of buildings—the huge number of small and simple facilities that to date have been left behind. The vast majority of commercial buildings are relatively small and lack any building automation or advanced maintenance system because the options are too complex. If managing a building becomes as easy as chatting with an AI and as motivating as leveling up in a game, even a small building with an overwhelmed part-time facility manager can take part.
The AI makes the action part of the habit loop incredibly simple (“just ask me”), and it can provide a tailored reward (the exact info or fix needed) much faster than a human could alone. Moreover, it can make the interaction feel engaging—even fun—because it’s a bit like talking to a super-knowledgeable colleague or playing around with a friendly oracle.
Jean-Simon Venne calls this approach “AI for UI,” meaning using artificial intelligence to reinvent the user interface. BrainBox AI’s interface, nicknamed “ARIA,” is designed so that “you’re basically going to talk to your screen rather than [use] mouse and keyboard,” Venne said. He likened it to JARVIS from the Iron Man films—essentially an AI butler for the facility.
Thanks to advances in voice recognition and large language models, the barriers that made this idea sci-fi a decade ago are largely gone. Venne noted that issues like AI hallucinations and slow response times, which were concerns initially, are being overcome with new techniques (for instance, grounding the AI in the building’s own data to keep it factual).
BrainBox AI had a head start here: “we built an agent with AWS before [agent] was a common term,” he said, describing how they trained their AI on a normalized dataset of building telemetry. They’ve since expanded it by ingesting equipment PDFs (manuals, specs) and comparing live sensor data to known performance curves. The result is an assistant that can answer complex operational questions by drawing on everything from design intent to maintenance history.
Venne gave an illustrative scenario of what’s now a beta feature: A large customer site in Louisiana calls in saying, “It’s too hot in here.” An operator getting that complaint might normally have to log into a building automation system (BAS), sift through trends for that zone, check the HVAC unit, maybe remote into a controls interface—a process that could take a couple of hours to pinpoint the issue (if it’s even found). In the AI-assisted world, that operator instead tells the AI, “Building 4, Zone 3 is too hot. What’s going on?”
The AI instantly parses this (understands the question and context—perhaps it knows the weather, the recent sensor readings, and that a work order was done on that unit last week). It then investigates across the data: it might pull the HVAC unit’s latest telemetry, cross-reference the maintenance log (e.g. a filter was changed last week), recall that the manufacturer manual suggests a certain error code for the symptoms, and even note that “oh, the optimal setpoint for this condition should be X and it’s currently set to Y.”
Within a couple of minutes, it returns an answer: “The cooling valve on AHU-7 appears to be stuck closed, causing the temperature rise in Zone 3. I recommend checking that valve; a temporary override has been applied to reduce supply air temperature.” In essence, it did the 2-hour investigation in 5 minutes and handed the tech a clear next step. As Venne put it, “You can do a lot more of these investigations than you can do today.” The technician’s capacity is supercharged—they can respond to many more complaints per day because the heavy diagnostic work is automated.
This intelligent assistant approach doesn’t just speed up individual tasks; it helps bridge the skills gap on the team. A junior tech with only a year on the job can now handle an issue that previously might have required a 20-year veteran’s intuition. The AI, having been “trained with a lot of HVAC documents” and past data, can infer the cause with “the same logic that a human would have followed,” according to Venne.
In other words, AI can act as a force multiplier for a shorthanded maintenance team. It won’t replace the humans (you still need someone to turn the wrench on that stuck valve or to verify the AI is correct), but it can guide less experienced staff to do more on their own and help senior team members accomplish more faster. Even things like suggesting replacement parts or optimal settings can be handled, reducing the cognitive load on the human and letting them focus on execution and higher value tasks.
75F’s Bob French shared a similar vision from the controls standpoint, which they’ve begun to realize in their product called “Saffron” (a nod to a helpful assistant). “You can talk to your building now,” French said. The intended user for Saffron isn’t the typical controls engineer—it’s “not a technical persona.” In fact, they talk about the concept of “the zero interface,” where an end user might not even know a building automation system is involved.
Imagine an office manager or a tenant who has access to a simple chat window or voice assistant for their building. If something’s wrong, they just state it in plain English: “The meeting room is too cold,” or “The lights on our floor are flickering.” Saffron takes that input, interprets it, and interfaces with the building’s systems behind the scenes. The user doesn’t need to learn any dashboard or terminology.
French gave a concrete example: “We’re calling for cooling, and the air coming out isn’t cool.” A non-technical person might report that observation. Traditionally, that would trigger a work order and likely a truck roll—a technician driving out to the site to see why the AC isn’t cooling. Often, the fix might be something as trivial as a setpoint that was set incorrectly or a system that needed a reset—tasks that don’t actually require a wrench, just knowledge of the control system.
With an AI assistant like Saffron, the system itself can recognize this scenario. It sees that the zone temperature is high despite a cooling command; it checks the equipment status and might discover that, say, the cooling setpoint was raised or a damper is closed. It can then either automatically correct the issue or guide the user: “I’ve adjusted the setting; give it a few minutes and it should start cooling. Let me know if it’s not better.” This avoids an unnecessary service call.
The maintenance team didn’t have to drop everything to drive over for a 30-second fix; they can reserve their time for truly complex issues or tasks that require hands-on work. This illustrates how AI can empower even tenants or non-engineers to solve minor building issues, which is another form of engagement—it expands who can interact with building systems beyond the typical facilities team.
Importantly, what 75F and others are doing here is reducing the training and expertise barrier to using building technology. “Prior to Saffron, [if] they needed end user training to implement a BAS… now, you can just give those commands to the AI and it does it for you,” French explained.
Historically, one reason smaller buildings (and their staff) never utilized advanced energy management or automation systems is that those systems were too complex. No one had the time (or inclination) to learn the user interface or interpret the data, so they’d give up. By hiding all that complexity behind a conversational AI, these vendors are essentially offering a “fluent translator” between humans and building machines. You don’t need to know which submenu has the air handler controls; you just say what outcome you want.
This means building technology can actually be used in the many places it wasn’t before. It’s a bit like how early computers were only used by specialists, until graphical interfaces and voice assistants made them accessible to everyone.
For years, the smart buildings industry has struggled with a paradox: there’s more and more tech available to run facilities better, but the people on-site often don’t use it.
Technicians, property managers, tenants… they’re all people, and people respond to feedback, challenge, and easy-to-use helpers. Giving a maintenance tech a pat on the back for a job well done (even if it’s just a digital high-five), or a challenge to beat their past performance, can turn a mundane workflow into something engaging.
Crucially, this convergence of gamification and AI could bring smart building benefits to the “long tail” of buildings—the huge number of small and simple facilities that to date have been left behind. The vast majority of commercial buildings are relatively small and lack any building automation or advanced maintenance system because the options are too complex. If managing a building becomes as easy as chatting with an AI and as motivating as leveling up in a game, even a small building with an overwhelmed part-time facility manager can take part.
The AI makes the action part of the habit loop incredibly simple (“just ask me”), and it can provide a tailored reward (the exact info or fix needed) much faster than a human could alone. Moreover, it can make the interaction feel engaging—even fun—because it’s a bit like talking to a super-knowledgeable colleague or playing around with a friendly oracle.
Jean-Simon Venne calls this approach “AI for UI,” meaning using artificial intelligence to reinvent the user interface. BrainBox AI’s interface, nicknamed “ARIA,” is designed so that “you’re basically going to talk to your screen rather than [use] mouse and keyboard,” Venne said. He likened it to JARVIS from the Iron Man films—essentially an AI butler for the facility.
Thanks to advances in voice recognition and large language models, the barriers that made this idea sci-fi a decade ago are largely gone. Venne noted that issues like AI hallucinations and slow response times, which were concerns initially, are being overcome with new techniques (for instance, grounding the AI in the building’s own data to keep it factual).
BrainBox AI had a head start here: “we built an agent with AWS before [agent] was a common term,” he said, describing how they trained their AI on a normalized dataset of building telemetry. They’ve since expanded it by ingesting equipment PDFs (manuals, specs) and comparing live sensor data to known performance curves. The result is an assistant that can answer complex operational questions by drawing on everything from design intent to maintenance history.
Venne gave an illustrative scenario of what’s now a beta feature: A large customer site in Louisiana calls in saying, “It’s too hot in here.” An operator getting that complaint might normally have to log into a building automation system (BAS), sift through trends for that zone, check the HVAC unit, maybe remote into a controls interface—a process that could take a couple of hours to pinpoint the issue (if it’s even found). In the AI-assisted world, that operator instead tells the AI, “Building 4, Zone 3 is too hot. What’s going on?”
The AI instantly parses this (understands the question and context—perhaps it knows the weather, the recent sensor readings, and that a work order was done on that unit last week). It then investigates across the data: it might pull the HVAC unit’s latest telemetry, cross-reference the maintenance log (e.g. a filter was changed last week), recall that the manufacturer manual suggests a certain error code for the symptoms, and even note that “oh, the optimal setpoint for this condition should be X and it’s currently set to Y.”
Within a couple of minutes, it returns an answer: “The cooling valve on AHU-7 appears to be stuck closed, causing the temperature rise in Zone 3. I recommend checking that valve; a temporary override has been applied to reduce supply air temperature.” In essence, it did the 2-hour investigation in 5 minutes and handed the tech a clear next step. As Venne put it, “You can do a lot more of these investigations than you can do today.” The technician’s capacity is supercharged—they can respond to many more complaints per day because the heavy diagnostic work is automated.
This intelligent assistant approach doesn’t just speed up individual tasks; it helps bridge the skills gap on the team. A junior tech with only a year on the job can now handle an issue that previously might have required a 20-year veteran’s intuition. The AI, having been “trained with a lot of HVAC documents” and past data, can infer the cause with “the same logic that a human would have followed,” according to Venne.
In other words, AI can act as a force multiplier for a shorthanded maintenance team. It won’t replace the humans (you still need someone to turn the wrench on that stuck valve or to verify the AI is correct), but it can guide less experienced staff to do more on their own and help senior team members accomplish more faster. Even things like suggesting replacement parts or optimal settings can be handled, reducing the cognitive load on the human and letting them focus on execution and higher value tasks.
75F’s Bob French shared a similar vision from the controls standpoint, which they’ve begun to realize in their product called “Saffron” (a nod to a helpful assistant). “You can talk to your building now,” French said. The intended user for Saffron isn’t the typical controls engineer—it’s “not a technical persona.” In fact, they talk about the concept of “the zero interface,” where an end user might not even know a building automation system is involved.
Imagine an office manager or a tenant who has access to a simple chat window or voice assistant for their building. If something’s wrong, they just state it in plain English: “The meeting room is too cold,” or “The lights on our floor are flickering.” Saffron takes that input, interprets it, and interfaces with the building’s systems behind the scenes. The user doesn’t need to learn any dashboard or terminology.
French gave a concrete example: “We’re calling for cooling, and the air coming out isn’t cool.” A non-technical person might report that observation. Traditionally, that would trigger a work order and likely a truck roll—a technician driving out to the site to see why the AC isn’t cooling. Often, the fix might be something as trivial as a setpoint that was set incorrectly or a system that needed a reset—tasks that don’t actually require a wrench, just knowledge of the control system.
With an AI assistant like Saffron, the system itself can recognize this scenario. It sees that the zone temperature is high despite a cooling command; it checks the equipment status and might discover that, say, the cooling setpoint was raised or a damper is closed. It can then either automatically correct the issue or guide the user: “I’ve adjusted the setting; give it a few minutes and it should start cooling. Let me know if it’s not better.” This avoids an unnecessary service call.
The maintenance team didn’t have to drop everything to drive over for a 30-second fix; they can reserve their time for truly complex issues or tasks that require hands-on work. This illustrates how AI can empower even tenants or non-engineers to solve minor building issues, which is another form of engagement—it expands who can interact with building systems beyond the typical facilities team.
Importantly, what 75F and others are doing here is reducing the training and expertise barrier to using building technology. “Prior to Saffron, [if] they needed end user training to implement a BAS… now, you can just give those commands to the AI and it does it for you,” French explained.
Historically, one reason smaller buildings (and their staff) never utilized advanced energy management or automation systems is that those systems were too complex. No one had the time (or inclination) to learn the user interface or interpret the data, so they’d give up. By hiding all that complexity behind a conversational AI, these vendors are essentially offering a “fluent translator” between humans and building machines. You don’t need to know which submenu has the air handler controls; you just say what outcome you want.
This means building technology can actually be used in the many places it wasn’t before. It’s a bit like how early computers were only used by specialists, until graphical interfaces and voice assistants made them accessible to everyone.
For years, the smart buildings industry has struggled with a paradox: there’s more and more tech available to run facilities better, but the people on-site often don’t use it.
Technicians, property managers, tenants… they’re all people, and people respond to feedback, challenge, and easy-to-use helpers. Giving a maintenance tech a pat on the back for a job well done (even if it’s just a digital high-five), or a challenge to beat their past performance, can turn a mundane workflow into something engaging.
Crucially, this convergence of gamification and AI could bring smart building benefits to the “long tail” of buildings—the huge number of small and simple facilities that to date have been left behind. The vast majority of commercial buildings are relatively small and lack any building automation or advanced maintenance system because the options are too complex. If managing a building becomes as easy as chatting with an AI and as motivating as leveling up in a game, even a small building with an overwhelmed part-time facility manager can take part.
Technology for facility maintenance has long struggled with engagement from the boots on the ground. Platforms like computerized maintenance management systems (CMMS) promise a “single source of truth” for work orders and equipment data, but too often, the people doing the actual work barely use them.
That’s a huge problem for building owners spending money on these tools—if technicians aren’t inputting data, the system can’t reflect reality. The bigger problem is what this gap represents: ineffective maintenance programs leading to missed work, undocumented issues, and reactive repairs.
One Accruent survey found that only 39% of businesses consistently use a CMMS to track maintenance tasks (the rest resort to spreadsheets or even pen and paper). This low adoption means many preventive tasks fall through the cracks, contributing to growing maintenance backlogs and costly breakdowns.
There are some good reasons for this poor engagement:
In short, today’s tools have not been very user-centric for the people on the ground, so those people ignore them—and the hoped-for efficiency and data insights never materialize.
Now that’s starting to change. Gamification, AI, and behavioral psychology principles influence users to return frequently, spend more time on the platform, and even enjoy doing so. Vendors aim to make facility software “sticky”—something folks habitually use, not dutifully avoid.
If that playbook sounds familiar, it’s because it’s the same engagement strategy that consumer tech (especially social media) has mastered over the last 20 years. Social media platforms leverage gamification tactics like:
These techniques, as described by author Nir Eyal’s popular Hook Model, create habit-forming loops of trigger → action → variable reward → investment. In other words, they cue you to engage, reward you in unpredictable ways that keep you guessing, and encourage you to put something of yourself into the system (time, effort, personal data) so you’re more invested in coming back.
While we can debate the merits of endlessly scrolling Instagram or chasing likes, there’s a growing case that some of these tactics can be repurposed to solve the long-standing engagement gap in building maintenance tech. What made Facebook addictive could make a maintenance app finally stick—helping buildings be better maintained, saving money, and improving occupant satisfaction (which can drive higher revenues for building owners through retention and reputation).
What’s especially exciting is the addition of modern AI into this formula. Natural language chat interfaces and AI assistants are supercharging the engagement strategy—and the capabilities of technicians and operators, too. A chat-based AI can draw users in with a conversational, almost human interaction, and then help them get things done faster than ever.
And this boost comes at a critical time: facility teams are being asked to do more with less, right as a generation of experienced technicians retires and a skilled labor shortage hits the industry. In the U.S., as Baby Boomers retire, 62% of companies are already struggling to fill skilled trade positions left behind. Maintenance departments everywhere feel understaffed and overextended. In this context, tools that actually engage the workforce—and extend their abilities with AI—aren’t just nice to have; they might be essential to keep buildings running reliably.
Facing these challenges, smart building technology vendors are taking pages from the gaming and social media handbook to better engage every persona that interacts with their systems. The key is recognizing that different users have different motivations.
As Jonathan Kroll, Co-founder and Chief Product Officer of Visitt, told us, “It’s gotta start with the user’s motivations, which are different for different personas.” A property manager, for example, already has a built-in incentive to address maintenance issues—it helps keep their tenants happy and makes them look good. But technicians and field operators? That’s a different story.
Kroll points out that technicians’ days are hectic and dynamic; stopping to fill out a digital form after fixing a toilet or AC unit can feel like just another task with no immediate benefit. But since this maintenance data is critical, the question becomes: how do you meet the technician where they’re at and give them a reason to use the system more?
One tactic is to make the software personally rewarding to use, beyond just the abstract promise that “data entry helps your company.” For example, Kroll mentioned the concept of an “inbox zero” achievement for work orders. This plays on a simple psychological urge—the satisfaction of completing a list and seeing it cleared. “We all want to clean our to-do list,” Kroll said.
When a technician resolves all their assigned work orders (or even just all the high-priority ones) for the day, the app might celebrate that with a big green checkmark, some on-screen confetti, or a congratulatory message. It’s a small dopamine hit, but it feels good—a moment of accomplishment instead of drudgery.
Another strategy is recognition and visibility. Technicians often express frustration that they “do a ton of work but nobody sees it.” In a busy facility, if an HVAC tech prevents a disaster or works through the night, tenants and bosses might never know—until something goes wrong. To address this, Visitt allows the end customer (e.g. the tenant) to provide quick feedback or ratings on a completed request. Kroll described how Visitt lets tenants give a simple 1–5 star review or a thumbs-up when their issue is resolved. That feedback goes straight to the technician’s profile. It’s also “celebrated” in the app, meaning the tech (and their team) might see a star count or a note like “Tenant X praised your response time!”
Over time, accumulating a high satisfaction rating becomes a point of pride. Management can even set a KPI like “maintain at least a 90% positive tenant rating” for the team, which gamifies the customer service aspect of maintenance. The important nuance: this isn’t about punitive metrics or public shaming; it’s about rewarding the frontline with recognition that historically they lacked.
Visually, the interfaces are also borrowing from consumer app design to make the experience more engaging. Instead of dry tables and forms, you’ll see friendly graphics when a task is completed or a progress gauge of how close you are to some target. These little UI touches act like the “likes” and badges of a maintenance app—small rewards or status signals that make the software less sterile.
Visitt created a “score” for building performance that updates based on maintenance activities and outstanding tasks. “Once you introduce the score, they want to get 100,” Kroll observed, noting how property managers and even executives got hooked on checking and improving their building’s score. It’s essentially a leadership board for buildings: if your property is sitting at a B grade (say 80/100) in maintenance health (maybe based on open issues, response times, etc.), that creates a natural drive to push it to an A.
Gamification isn’t just for the in-house staff—it’s also being used to engage service vendors and contractors who work on buildings. Bob French of 75F gave an example of how they’re applying game design to the commissioning and maintenance of HVAC control systems. 75F recently launched a feature called “Easy Street” which, as French explains, “gives a site installation score and gamifies the commissioning process.”
Here’s how it works: when a contractor installs 75F’s HVAC controls in a building, they run the Easy Street automated checkout. The system then analyzes whether all controllers and sensors are installed correctly and calibrated. Instead of just spitting out a dry report of issues, it produces a score—essentially rating the quality of the install. If you, the installer, got an 86 out of 100, you know there’s room for improvement. The tool also generates a punch list of what needs fixing (maybe a sensor offline or a config not optimized).
They can address the punch list items and run it again to get closer to that coveted 100. 75F envisions technicians really taking to this, even to the point that “if someone gets a good score then they will go on LinkedIn and brag about it.” In other words, turning a perfect installation into a badge of honor in the professional community. The score gives immediate feedback and a sense of achievement for doing a thorough job, which is much more motivating than a long checklist with no clear indicator of success.
That same scoring concept continues after installation as well. A facility manager can periodically run the Easy Street diagnostic on their building (whether monthly or quarterly) to see if performance is drifting. They’ll get a score each time, and perhaps more importantly, see the trend. Maybe last quarter the building was a 92, and now it’s down to 88—time to investigate and bring it back up.
This creates an ongoing game of keeping the building performing at its best. The historical log of scores taps into our natural desire to improve our “personal best” and not let it drop. French expects facility teams to take pride in maintaining high scores, which in practice means proactively fixing issues the software identifies.
Jean-Simon Venne of BrainBox AI agrees: make operating a building feel like playing a game. Instead of discrete rewards like badges or points, he says “it’s about performance… Can I get closer to 100%?” Venne described how any building system (say a chiller plant) has an optimal performance curve—essentially the most efficient it could be running under current conditions. If you give the operator a real-time metric of how close they are to the optimal, it becomes a challenge to push it closer.
Crucially, Venne also highlighted why current tools struggle to engage: chaotic UI and information overload. Traditional building automation systems and cloud applications often bombard users with data—dozens of tabs, graphs, and raw values for every piece of equipment. “The problem… is that traditional software UIs have so many tabs. The more data you have, the more tabs you have,” he noted.
A veteran operator might navigate that, but your average maintenance tech or property manager just finds it bewildering. “You just want to know, is my building okay or not? Should I send a truck or not?” Users need simple answers to important questions, not a sea of charts.
Gamifying operations means streamlining the experience so it’s easy to consume (games usually teach you gradually and keep the interface intuitive). If a tool can tell an operator “All systems go—you’re at 95% optimal” or “Alert: you’re dropping—here’s what to do”, that concise insight keeps them engaged far better than dumping a PDF report on their lap.
The next evolution of this trend is weaving AI and natural language interfaces into building operations, turning the whole experience even more interactive and personalized. If gamification provides the motivation to use the tool, AI is providing a new means to use it—one that can drastically reduce friction and boost productivity.
An AI assistant changes the game (pun intended) by making the software feel less like software at all. Instead of rigid menus and forms, users get a conversation partner that can understand requests and guide them to solutions. This adds a new layer to the habit-forming loop: the system isn’t just reactive, it actively engages with you, almost like another teammate.
Think of how people started engaging with ChatGPT and similar AI assistants in the past couple years. Millions of users found themselves coming back again and again to these bots to ask another question, try another idea—partly because it’s so easy (just type what you want, no training needed), and partly because the responses are varied and sometimes surprisingly helpful. It’s the Hook Model in action: you have a question (trigger), you prompt the AI (action), you get an answer that’s sometimes great, sometimes mediocre (variable reward), and then you refine your approach or give feedback (investment), which makes the next interaction a bit better.
It can become addictive in a useful way. In fact, ChatGPT became the fastest-growing app in history, reaching 100 million users in just 2 months—a testament to how compelling an AI-driven interface can be when it works. Now apply that concept to a building management context: instead of digging through manuals or ignoring complex software, what if a building operator could just ask an AI assistant “Hey, why is it so hot on the 3rd floor today?” and get a credible answer? That’s where we’re headed, and it builds on all the gamification groundwork described earlier.
Log in or join Nexus Pro to read on... we'll cover exactly how AI chat interfaces work to enable this and what it means for FM.
Head over to Nexus Connect and see what’s new in the community. Don’t forget to check out the latest member-only events.
Go to Nexus ConnectJoin Nexus Pro and get full access including invite-only member gatherings, access to the community chatroom Nexus Connect, networking opportunities, and deep dive essays.
Sign Up
Nexus Labs exists to move the smart buildings industry forward, without the fluff.
On Wednesday, we send out the renowned (and free) Nexus Newsletter.
This is a great piece!
I agree.