Webinar: Building an AI-First Team

Most software companies have plenty of people using AI on their own, but few have figured out how to build a team that runs on it. That gap, between one person quietly getting more efficient and an entire organization operating differently, was the focus of Mainsail’s recent webinar, “Building an AI-First Team.”

Nick Olsen, Head of AI Innovation at Mainsail, sat down with Bill Simpson, CTO-in-Residence at Mainsail, and Michael Boord, President & COO at ServiceCore, a vertical software company in Mainsail’s portfolio, for a live webinar on what it takes to move a team from scattered AI experiments to daily, organization-wide use. Watch the recording here, or read on below for the six principles they covered.

1. Model It: You Have to Get Caught Using AI

For a long time, people would sand off anything that looked AI-written and spend more energy concealing the tool than the work itself took. That instinct has to go. If you want your team to treat AI as a default instead of a shortcut to hide, you have to let them see you using it.

At ServiceCore, Michael treats this as a change management problem first. When someone brings him a question, his first move is to open Claude and work the problem live with them, instead of reaching for the way he would have solved it a year ago. He also actively shares the skills, prompts, and examples that are working for him, so the team has a running supply of proof that this is simply how the company operates now.

Bill frames it as psychological safety. Admitting what you don’t know, in front of your team, is what gives everyone else permission to be a beginner too. His advice to CTOs: pick a hard problem, spend the weekend learning it with Claude, and come back Monday and tell the team exactly how you did it.

Try this

Find one way this week to let your team see you use AI that you haven’t shown them before.
In your next one-on-one, open the tool live and work a problem in front of the person you’re talking to.

2. Be a Great AI-Using Teammate

The flip side of visible AI use is volume. AI makes it easy to produce far more content than anyone asked for, and the result, if left unchecked, is bloated documents nobody (even the “author”) reads.

Our panel’s advice is to hold people accountable for what they submit. If someone sends an 18-page document, read it and ask direct questions about the vague bullet point or the AI-isms buried inside it. “Oh, that was Gemini” is not an answer; it’s the coaching moment. A document shouldn’t run 15 pages unless the content truly demands it, and teams should build skills that train AI tools to write in their own voice instead of a generic one. Make this a habit by pulling two weeks of a team’s output, read it closely, and ask people to walk you through what they shipped in a one-on-one. If they can’t explain it, that’s the coaching conversation, and a reminder that the job is to be the discerner of AI’s output, not just the forwarder of it.

Try this

Build your own AI “voice” skill so AI-assisted work sounds like you, not like generic AI output.
Set a real ceiling on document length and ask pointed questions about anything that reads like it wasn’t reviewed.

3. Put Agents on the Org Chart

If you’d put a new hire on your org chart, hold your agents to the same bar. That means a defined scope, a standard you measure against, and a willingness to invest further, or cut it loose, based on whether it’s actually productive.

Before a team requests a new hire, have them show that an AI-enabled workflow can’t do the job, typically backed by a 30-to-60 day measurement period with real numbers, not a single afternoon spent poking at a chatbot. The bar for approving new headcount will go up considerably as a result.

Try this

Before approving your next headcount request, require proof, backed by a real measurement period, that AI can’t do the job.
Look at your open or upcoming roles and ask which parts of the job are really “builder” work versus “doer” work an agent could absorb.

4. Set the Bar

If your open job descriptions now expect AI fluency from candidates, your current team needs to hear that the same bar applies to them. That’s an uncomfortable conversation to have out loud, which is exactly why it has to be explicit rather than implied.

Michael was clear to his teams that the company would invest heavily in tooling, training, and departmental champions so employees had every resource to build these skills, and in return, employees needed to lean in to reach a baseline fluency level by a set date.

Try this

Say out loud, to your current team, the AI expectations you’d already put in a new job posting.
Pair the expectation with enablement, tooling, training, and champions, so the ask feels like an opportunity, not a threat.

5. Address the Fear

Mainsail’s panel didn’t soften this one: AI is going to change your job and pretending otherwise doesn’t help anyone. The more useful move is naming that directly and giving people a way to help rewrite what their role looks like next.

Bill pushed back on calling it just “fear.” For people who spent years building one specialized skill, what’s underneath is often grief, a real sense of loss, not simply anxiety. He’s watched it play out generationally: people early in their careers adapt fastest because they have less invested in the old way of working, and people several career pivots in treat AI as just the next language to learn. It’s the people in the middle, who spent the last several years mastering one specific skill set, who feel it the hardest.

Michael describes this moment as one of the rare platform shifts that comes once every fifteen to twenty years, a chance for someone to get a step function jump in their career instead of the usual slow climb, if they choose to lean in.

Try this

Have this conversation one person at a time, not in a company-wide meeting.
Name what someone might be feeling, including grief, not just fear, before you try to talk them out of it.

6. Build the Ladder

Telling people they have a career opportunity is a platitude unless you show them the actual path. ServiceCore’s answer was a five-level AI fluency rubric (AI Aware, AI Developing, AI Integrated, AI Amplified, and the aspirational AI Transformed), built separately for every team in the company, defining specific behaviors, tools, workflows, and measurable signals at each level.

The idea is that, regardless of where someone starts, everyone can reach AI Integrated, meaning daily use woven into core workflows, by the end of the year. Michael built the rubric quickly by creating an interview skill that department heads could run in their AI tool of choice, which walked them through defining behaviors and workflows for their own team. A rubric that once would have taken days or weeks came together in about 45 minutes to an hour.

Try this

Build a simple, team-specific fluency rubric, even a rough one, instead of a single company-wide AI mandate.
Set one minimum level everyone is expected to reach, and a real date to reach it by.

The Bottom Line

For a software company, building an AI-first team is less about picking the right tool and more about leadership. Modeling visible use, holding your own team accountable for what they ship, treating agents like hires, setting an explicit bar, naming the fear directly, and building a real ladder to climb are all versions of the same job: giving people permission and a path, not just a tool.

Asked for one piece of advice for becoming an AI-first company, the panel’s answers converged on the same idea: be explicit about what you expect from your team and be the one setting the example rather than just handing down the mandate. Don’t wait for the perfect plan. Start today, wherever you are, and take the next step.

This content piece has been prepared solely for informational purposes. The content piece does not constitute an offer to sell or the solicitation of an offer to purchase any security. The information in this content piece is not presented with a view to providing investment advice with respect to any security, or making any claim as to the past, current or future performance thereof, and Mainsail Management Company, LLC (“Mainsail” or “Mainsail Partners”) expressly disclaims the use of this content piece for such purposes.

The information herein is based on the webinar guests’ opinions and views and there can be no assurance other third-party analyses would reach the same conclusions as those provided herein. The information herein is not and may not be relied on in any manner as, legal, tax, business or investment advice.

Third-party images, logos, and references included herein are provided for illustrative purposes only. Inclusion of such images, logos, and references does not imply affiliation with or endorsement for or by such firms or businesses.

Certain information contained in this content piece has been obtained from published and non‐published sources prepared by other parties, which in certain cases have not been updated through the date hereof. While such information is believed to be reliable for the purposes of this content piece, neither Mainsail nor the contributors to this piece assume any responsibility for the accuracy or completeness of such information and such information has not been independently verified by any of them. The content piece will not be updated or otherwise revised to reflect information that subsequently becomes available, or circumstances existing or changes occurring after the date hereof, or for any other reason.

Certain information contained herein constitutes “forward-looking statements,” which can be identified by the use of terms such as “may,” “will,” “should,” “could,” “would,” “predicts,” “potential,” “continue,” “expects,” “anticipates,” “projects,” “future,” “targets,” “intends,” “plans,” “believes,” “estimates” (or the negatives thereof) or other variations thereon or comparable terminology. Forward looking statements are subject to a number of risks and uncertainties, which are beyond the control of Mainsail. Actual results, performance, prospects or opportunities could differ materially from those expressed in or implied by the forward-looking statements. Additional risks of which Mainsail is not currently aware also could cause actual results to differ. In light of these risks, uncertainties and assumptions, you should not place undue reliance on any forward-looking statements. The forward-looking events discussed in this content piece may not occur. Mainsail undertakes no obligation to update or revise any forward-looking statements, whether as a result of new information, future events or otherwise.

No representation, warranty or undertaking, express or implied, is given as to the accuracy or completeness of the information or opinions contained in the enclosed materials by Mainsail and no liability is accepted by such persons for the accuracy or completeness of any such information or opinions. For additional important disclosures, please click here.