(Image credit: IIR.)

In February, ALKEME Insurance (Ladera Ranch, Calif.) put its name on one of college basketball’s best-known venues. The 15-year, $27.7 million agreement renamed the University of Arizona’s McKale Center as McKale Center at ALKEME Arena—a conspicuous sign of the brokerage’s rapid emergence since its founding in 2020.

ALKEME now ranks as the 16th-largest U.S. P&C agency, with more than 1,500 employees and 90 locations. Yet the company’s ambition involves more than accumulating agencies and building a national brand, according to Curtis Barton, CEO, ALKEME. Barton describes ALKEME as an operator-led brokerage designed to integrate its acquisitions, give agency leaders greater resources and return producers’ attention to clients and growth.

“We’re not trying to be a traditional insurance brokerage,” Barton says. “We’re doing it differently. We’re putting emphasis in areas where our competition is not right now, or where they’re dabbling and we’re taking it very seriously.”

Technology has become central to that proposition. ALKEME is consolidating acquired agencies onto common systems and data while building proprietary applications around the work of producers and account managers. Its objective is not to introduce AI as a separate tool employees must learn to use, but to embed it selectively within the workflows where it can identify opportunities, eliminate repetitive work and reduce the number of systems employees must navigate.

“We try to bake it into workflow,” says Ryan Deeds, Head of AI, ALKEME. “It’s not, ‘Here’s AI—go use it.’ It’s, ‘Here’s where it can help you, and here’s how you interface with it.’”

An Operator-Led Brokerage

ALKEME began in 2018 as a group or cluster before taking its current form with private-equity backing in 2020. Barton had previously grown a retail brokerage to approximately $11.5 million in revenue. When he considered how to double its size, he began talking with brokerage consolidators.

What he encountered convinced him that there was room for a different model.

“They really weren’t insurance people,” Barton recalls. “They were more banking people doing financial engineering. At the time, it was more about how fast you could deploy capital. Who cared about integration and migration of your data? Who cared about data governance? Nor was organic growth really being discussed.”

Barton believed that acquiring agencies without integrating them would restrict the resulting organization’s ability to improve operations or make effective use of its combined data. ALKEME therefore began with the intent to become one company rather than a cluster or loosely connected association.

The brokerage acquires agencies specializing in particular industries or types of risk, with sellers generally receiving a combination of cash and ALKEME equity. Local operators gain carrier access, centralized services, technology, data capabilities and other resources that would be difficult to assemble independently.

The proposition, Barton says, is to relieve agency owners of functions that distract them from production and client relationships.

“Every guy or woman who runs one of these organizations originally started as a salesperson,” Barton says. “They didn’t start as an HR person. They get caught in the quagmire of how to run the business, continue to grow it and look at technology. We can take that off their backs and get them focused on selling.”

ALKEME nevertheless expects acquired agencies to join a common enterprise. “Every deal that we buy, we integrate,” Barton says. “You take our name, and you take our system.”

Integration Before Innovation

For most retail agencies joining ALKEME, that system is Applied Epic from Applied Systems (Chicago). Deeds says approximately 90 percent of acquisitions migrate to Epic within seven months, although timing varies according to an agency’s size, systems and data quality.

A well-run agency with orderly systems could be integrated in as little as a month, Deeds says. He cites a $15 million-revenue agency that ALKEME migrated from Vertafore Sagitta to Epic within five months, including the associated data conversion and employee training.

ALKEME makes exceptions when an agency operates effectively with technology suited to a specialized business. An employee-benefits agency with highly developed Salesforce workflows, for example, may not be an immediate candidate for conversion. ALKEME has also created integrations that allow agencies to retain applications such as AgencyZoom or HubSpot while connecting them to Epic.

The objective is not uniformity for its own sake. It is to bring the information required to run the enterprise into a common environment. ALKEME consolidates policy, client, contact and activity data along with HR, payroll, email, Microsoft Teams and other Microsoft 365 information. Its central environment uses a MongoDB document-based data architecture.

Integration begins even before an acquisition closes. ALKEME’s internally developed M&A Portal replaces a collection process that previously involved multiple forms, systems and handoffs. The portal tracks acquisition prospects, assigns tasks, gathers business information and begins assembling data about the agency.

New agencies subsequently enter Producer Portal, an internally developed operating environment that gives producers a common view of goals, production, leads and other business information.

“When I got here, nobody spoke the same language,” Deeds says. “You couldn’t tell me which producers were with which agencies. You couldn’t tell me what their goals were. Now all of that is done.”

Producer Portal was initially launched in July 2025 and has approximately 500 potential users. Deeds says about 360 use it during a typical month, with approximately 110 daily users. Account managers are beginning to enter the environment as ALKEME adds service-oriented capabilities.

The technology foundation provides more than consolidated reporting, Deeds emphasizes. It gives ALKEME common definitions, repeatable processes and the ability to build applications that work across the organization.

“You can’t really do AI until you have good standard operating procedures,” he says. “We might have had 38 different ways of doing a personal-lines renewal. Which is the golden path? How do we determine that? Who has the right to make that decision? The human element takes much longer than the technology today.”

Finding Repetitive Work

One of ALKEME’s clearest automation examples involves an unglamorous but pervasive activity: renaming electronic documents delivered by carriers.

Carrier documents arriving through Ivans may carry names that do not adequately identify their contents. An employee previously had to open each document, determine its type, carrier, policy and effective dates, and rename it accordingly. ALKEME calculated that employees performed the task approximately 160,000 times over a 12-month period, with each instance potentially requiring five to 10 minutes.

ALKEME examined approximately 290 activities within its agency management environment, evaluating the time and complexity involved in each. Document renaming emerged as an attractive automation candidate because the company possessed abundant historical data and could establish a high confidence threshold.

The resulting capability now renames approximately 70 percent of incoming documents automatically, according to Deeds. Items that do not meet the required confidence threshold enter an exception queue for human review.

The example illustrates an important distinction in ALKEME’s AI strategy. AI assisted the company in building the application and may help classify uncertain documents, but much of the resulting process runs through deterministic software. The objective is not to insert generative AI everywhere, but to use the appropriate technology to remove a recurring operational burden.

“So much of what we’ve built has been about getting us onto the same operating page,” Deeds says. “Before, we were spreadsheet-run. We’re not running critical processes out of spreadsheets anymore. Everything is in systems.”

Turning Data Into Growth Opportunities

ALKEME is also applying its data and internally developed technology to organic growth. SalesRoom, the company’s CRM and sales-enablement environment, combines prospecting data, communications, account-rounding opportunities and campaign functionality.

The system analyzes available account information to identify clients that may lack relevant coverage. Producers periodically receive a selection of potential account-rounding opportunities and decide whether the recommendation is appropriate. If the producer elects to proceed, the system proposes an email for review and approval.

The communication can be adapted to the producer’s own style. With the employee’s authorization, the system analyzes a sample of previously sent emails to identify common phrases, tone and writing patterns. It then uses that profile when preparing proposed communications.

The employee remains responsible for deciding whether the coverage opportunity is valid and whether the message should be sent. During the first weeks of one newly deployed rounding process, the system generated more than 150 client outreach actions that otherwise would not have occurred, according to Deeds.

ALKEME is also using the environment to conduct larger campaigns. At the time of the interview, the brokerage was preparing cyber-related outreach to approximately 3,100 clients. The messages were to be delivered through producers’ accounts after they had reviewed and approved samples. Results from that campaign were not yet available.

The same data can generate a summary of a client relationship by drawing upon policy, activity and communication records. ALKEME’s system checks the generated account summary against source materials before presenting it to the employee.

Deeds calls ALKEME’s operating principle “human at the edge”: technology gathers information, performs repetitive work and proposes an action, while the insurance professional verifies the result at the point where judgment is required.

“We don’t allow AI to do much from the insurance perspective by itself,” he says. “We identify opportunities with AI, and sometimes they’re wrong. Everything that gets sent out is looked at by a human.”

ALKEME also uses an additional AI review mechanism to examine outbound campaign messages. If the recipient name, company, content or intended audience appears inconsistent, the system can hold the message as an exception. AI in that instance serves as a safeguard on human action rather than as the original decision-maker.

The company currently avoids having AI provide autonomous coverage advice, Deeds stresses. Its prospecting communications focus on why a business might want to work with ALKEME rather than purporting to make an independent insurance recommendation.

Rapid Development Around the User

ALKEME’s internally developed technology reflects an operating philosophy that begins with employees’ problems. Users can submit suggestions inside Producer Portal, after which Deeds’ team may build and release a solution for testing.

In one instance, an employee requested a vehicle identification number tool. Deeds’ team located an appropriate application programming interface and produced a working page within two hours, making the capability available to approximately 60 employees.

“Our goal has always been to build software extremely fast, get it to the user and have them use it and iterate over it,” Deeds says. “If they use it, we get a really good product because they iterate over it.”

That model also helps ALKEME preserve valuable differences among acquired agencies. The brokerage standardizes the data and processes it needs to operate cohesively, but it can retain specialized systems or build accommodations when a local agency has developed an effective way to serve a particular market.

Adoption, rather than the novelty of the technology, is the ultimate test, Deeds suggests.

“The person coming into your product has to have it make their day better—make them feel safer, more organized or faster,” he says. “If it doesn’t do that, then it’s not going to work.”

Toward an Agentic Operating Environment

ALKEME’s most ambitious development effort is Liquid, an internally built agentic interface intended to let employees work across Epic, Outlook, Teams, Word and Excel without repeatedly moving among applications.

Liquid remains in testing, with production deployment anticipated in early 2027. In Deeds’ description, an account manager could ask the system to identify important incoming emails, assemble information needed for a response, propose language and send the approved message. The employee could then direct an agent to address a group of agency-management-system audits, review the proposed changes and authorize the system to make them.

The approach extends “human at the edge” from individual AI-assisted tasks into a broader working environment. Agents perform the navigation and preparation; employees retain authority over consequential actions.

ALKEME is beginning with account managers because much of its initial development concentrated on production. Deeds’ team has identified approximately 10 recurring account-management activities that consume substantial time and is using them to test whether the model can reliably remove work from employees’ desks.

Liquid also exemplifies Barton’s belief that companies of sufficient scale will increasingly develop software suited to their own operating models.

“Any company of scale will become a semi-software developer of its own,” Barton says. “To what degree is the question.”

For ALKEME, proprietary technology remains a means of strengthening the brokerage rather than a separate product strategy. Its purpose is to improve visibility, decision-making, efficiency and producer support. Over time, however, Barton believes ALKEME’s technology and data capabilities may become an important part of the company’s competitive intellectual property.

The deeper differentiator, Deeds argues, is a culture willing to consider what has recently become possible and take controlled risks to explore it.

“People get stuck on what they think is currently possible,” he says. “Our ability to see what might be possible, redefine it and take controlled risks to navigate those possibilities—that, to me, is a core differentiator.”

Applied Launches Agentic Email-to-Quote Channel