Image credit: Pexels
From petition drafting and property management to sales, operations, and content strategy, business leaders are using AI to cut repetitive work while keeping human judgment at the center.
Artificial intelligence has moved beyond the buzzword stage and into day-to-day business operations. Across industries including immigration law, property management, business services, social media, and nutrition, leaders are realizing that the strongest value lies less in replacing people and more in expanding what teams can accomplish.
Examples from businesses show a common pattern: AI works best when it takes on repetitive, data-heavy, or time-consuming tasks, leaving people more room for judgment, relationships, and strategic decisions.
Cutting Petition Drafting Time by 90%
For Mahmudul Hasan, founder and CEO of Clarvo, this value is particularly clear in immigration law. Clarvo’s AI automates approximately 70% of the initial drafting of EB-1A and EB-2 NIW petitions, cutting work that previously required 20 to 30 hours to roughly two or three hours.
This shift lets lawyers and law clerks spend more time reviewing and strengthening petitions rather than starting each document from scratch.
Mahmudul Hasan explained: “If I have the 70% work done in two hours, or less than three hours, I can spend the remaining 17 hours to improve the quality of the petition. To make sure that this is not just AI work—we are not just submitting the production by the AI. The lawyer’s input and insights are there.”
Hasan, however, sees human oversight as essential because a seemingly small omission can affect a case.
“This is where the judgment of a human lawyer matters, because if you give it to the AI, it might overlook that specific keyword or that specific phrase. The lawyer might find it damaging to the case.”
The risk also extends to the other side of the process, Hasan said.
“USCIS is also using AI at the same time. That is another aspect of risk, because sometimes we don’t know what kind of keywords they’re looking for. They might change those keywords from time to time, depending on priorities, depending on policies.”
Turning Property Calls Into Sales Opportunities
At Brickwise AI, AI is being applied to streamline repetitive property-management work. The company, which has raised $3.5 million after graduating from Y Combinator, developed its own LLM to handle tenant calls, vendor coordination, invoicing, compliance, and reporting.
Andreis Bergeron, head of GTM and partnerships at Brickwise AI, said the system also supports the company’s sales operation by refining ideal client profiles and improving outbound communication.
“We’ve created our own LLM for property managers, basically handling all the low-level, repetitive tasks: inbound tenant calls, someone calling in about renting a property, rent being late, compliance and reporting. All of that is handled by an LLM, making decisions and sitting on top of their current tech stack.”
The company has also seen its close rate rise from roughly 15% to more than 25%, while its reply rates have moved beyond the industry’s saturated 0.3% baseline.
“You don’t realize how much is slipping through the cracks. I’d say you probably move from like a 15% close ratio to like a 25-plus percent close ratio. You have the full conversation recorded and a customized email sent out answering the questions from the meeting; that’s what actually moves the sale,” Bergeron added.
For Bergeron, those gains reinforce a broader principle.
“The question you should be asking is: what is the thing I can do better than AI, and what are the things I can’t? AI is really good at filtering through large sets of obscure data. What we’re good at as humans is having our perspective; you don’t want to get rid of that.”
Scaling Operations Without Simply Adding People
Makena Finger Zannini, founder and CEO of The Boutique COO, has taken a different route. Rather than treating growth as a reason to increase headcount, she focused on eliminating hundreds of small manual handoffs, from email drafting to onboarding and coordination.
Those incremental changes allowed the firm to reduce its team from 10 full-time employees to 1.5 while increasing capacity.
According to Makena Finger Zannini: “It’s all these little tiny things that are easy to get missed by humans, and they just add up to a lot of time. It’s these little improvements, chipping away at it instead of trying to pull things up from the root, that’ve allowed us to go from 10 full-time employees to one and a half, while actually doing more.”
The time saved hasn’t simply disappeared from the business. Zannini has redirected it toward activities that depend heavily on personal interaction.
“What I’ve been spending more time on is relationship management, business development, sales. People want more analog experiences: in-person meetings, events. AI has actually allowed me to do more of that. I’m going to more events, having more conversations. It’s allowed us to meet this moment of people wanting that analog connection.”
That experience has also shaped what Zannini chooses not to automate.
“AI-generated content, AI-generated marketing, AI-generated copywriting—I get the temptation, but it just doesn’t work. We run marketing for hundreds of small businesses, and everyone who wants to post this AI-generated content, it’s not getting engagement. For me, it’s not even a normative thing, it’s just purely that it doesn’t work. I’m telling you, you’re not going to get the results.”
Using AI as a Thinking Partner, Not a Substitute
At Yuvoice, CEO Isvari Maranwe uses AI for research, branding, and algorithm development while maintaining strict limits around its role. The social platform itself is designed to reward real-world impact through a Karma Point system and is funded by a social commerce marketplace rather than advertising.
Maranwe views AI as useful when it challenges assumptions and accelerates research.
“When you use AI as a tool, as a way to bounce ideas off of or do quick research, it’s kind of like another thinking partner. One of the best ways to do this is to train it with system-level prompts to challenge everything you say and to be really clear about when it doesn’t know something. Anything you’re fundamentally going to put out, you still research and back from first principles after.”
But Maranwe argues that human oversight cannot be treated as optional.
“It’s really important to have a human in the loop, for basic ethics, but also for your own company’s sake. There are already instances where AI systems competing against each other are inserting malware, corrupting systems to achieve the one goal you told them, because you didn’t tell them not to, and they didn’t know. Just put a human in the loop.”
Her warning focuses on how quickly AI systems can move from assistance to action when given broad access.
“The most dangerous misconception is that people underestimate how powerful and capable AI is. I worked with a company that basically destroyed their systemic data—they gave AI direct access to clean things up, it erased the backups too, and couldn’t be retrieved. Business leaders need to understand the vast difference between using stable, existing AI and deploying frontier models. That distinction is critical.”
From Performance Reporting to AI Engine Optimization
For Chris Manderino, founder and CEO of LyfeFuel, AI became a practical response to inefficient reporting and underperforming agency work. The company uses Victor to pull data from Shopify, Amazon, Meta, and Google into a unified Slack dashboard, eliminating manual reporting and improving performance visibility.
Manderino describes the tool as a broad operational assistant.
“Victor is almost like an executive assistant across the whole business; it’s got the superpower of a hundred VAs. We went through an audit on everything and what we’ve been able to uncover and improve has been a game changer. There are things we used to pay agency contractors for where we were getting half-assed work done. Now Victor does it, and it’s not only saved us money on agency fees, but also improved performance.”
Manderino also uses a “Board of Five” prompt to challenge major decisions by simulating different perspectives.
“I call it the Board of Five. There are five different mindsets: the contrarian, the optimist, the pessimist, whatever. For bigger decisions, you run it through the board and get the full breakdown. Four out of five concur, but the one pessimist says, ‘here’s the things to watch out for.’ It’s how I stress-test a decision before committing.”
His next focus is AI Engine Optimization, or AEO, as consumers increasingly use AI agents to find recommendations.
“Instead of going on Google and digging through affiliate links, now you’ve got a multi-threaded prompt when you’re searching for a solution. We can dig into the customer psychology, understand what queries they’re looking for from a prompt engineering standpoint, and ensure we have things throughout the website not built for humans to read, but for AI agents to read, so we get discovered and put in front of the right people. That’s what they call AEO, AI Engine Optimization.”
Final Thoughts
Across these businesses, the applications differ, but the underlying lesson stays the same. AI creates its clearest value when it absorbs repetitive work, processes large amounts of information, and removes operational bottlenecks. The people using it still provide the judgment, relationships, and perspective that determine how that technology is applied.
For business leaders, the question is becoming less about whether to use AI and more about where it can make a measurable difference and where human involvement remains indispensable.
