More Agents Won’t Fix Unpredictable Projects: Why We Built a Superagent Instead
Walk into any professional services firm right now and you’ll find AI everywhere. Resourcing agents. Project management assistants. Forecasting tools. Finance copilots. Each one solving a specific problem for a specific team, and each one completely unaware of what the others are doing.
The industry calls this AI progress. Kantata calls it the problem. And it’s making unpredictable projects worse, not better.
On a recent episode of the Professional Services Pursuit podcast, Kantata CEO Michael Speranza and Chief Product Strategy Officer Sarah Edwards made the case for why professional services firms need something fundamentally different — and introduced what Kantata built in response: the Expertise Agent, the first AI superagent designed specifically for professional services.
The Problem Isn’t a Lack of AI
Professional services is unlike almost any other industry in how deeply interconnected its operations are. A single resourcing decision doesn’t live in isolation. It touches delivery timelines, client outcomes, revenue recognition, and team capacity all at once. Change one thing and you’ve changed everything downstream.
That’s what makes generic AI tools such a poor fit. They’re built to answer questions fast — but fast isn’t the same as accurate when the answer depends on context that spans your entire business. A tool that optimizes one team’s workflow without understanding the ripple effects across the rest of the organization isn’t solving the problem.
It’s surfacing a different version of it.
Sarah sees this play out in conversations with customers every day: “I think they are implementing AI in silos or they’re implementing point solutions. But the true opportunity they face in the services business is: how do I operationally connect all of that and really understand how all those decisions are connected?”
The more tools a firm layers on, the more fragmented the picture becomes. What looks like an AI-enabled operation is often just a collection of agents pointing in different directions, with no shared understanding of how the business actually runs.
Fast Doesn’t Always Mean Accurate
There’s a false confidence that comes with deploying AI quickly. Sure, teams feel productive and decisions get made faster. But speed without context produces a specific kind of risk: decisions that feel informed but aren’t.
Generic AI compounds this because it lacks vertical expertise and the ability to continuously learn from your data. It gives you the best horizontal answer it can — which is rarely the best answer for a professional services firm navigating the complexities of a live portfolio. Michael puts it plainly: “What we’ve seen is folks really running for those false positive answers. And not that they’re bad decisions necessarily, but are they the best decision possible for their business? And I think the risk that you run there [is that] everyone’s going to run towards mediocrity. [They] have these fast, mediocre decisions. And I think that is a huge risk that everybody should pay attention to.”
“The firms that are really going to flourish are the ones that take a step back, look and understand what the right long-term technical differentiation strategy is going to be. If it’s available to everybody, is it really differentiating at the end of the day? The answer to that is no.”
– Michael Speranza, CEO
One Agent that Sees the Whole Business
Kantata’s answer to AI sprawl wasn’t to build a better version of the same thing. It was to rethink the category entirely.
The result? The Expertise Agent.
The Expertise Agent is a superagent, one that understands the full operational context of a professional services firm and can take action across projects, resources, financials, and systems simultaneously. This isn’t a tool scoped to a single role or workflow, but something extensible to every person in the business, from consultants to practice leaders to the executive team.
The distinction matters because the alternative – building narrowly defined agents for specific roles – only serves a fraction of the workforce. It doesn’t meet the needs of what clients are actually asking for. They don’t want just task automation or just a smarter reporting layer; they want an intelligent partner that helps the entire organization become a learning organization, one that captures institutional knowledge, applies it across every engagement, and gets better over time.
“What we have built is something that is extensible to everybody,” Michael explains. “You don’t have to know what your role is. You don’t have to know what your job is. You can log in, interact with our superagent, and it will help you do your job. It knows who you are, it knows what your role is, and it leverages all the information to understand what question you’re asking it — and how it can help you. It’s not imagining the system in a series of lanes that don’t connect. It knows that all of these lanes connect every single day.”
Predicting What You Couldn’t See Before
When you have a system with full operational context, the kind of predictability that PS firms have been chasing starts to become real. Delivery risks surface before they hit the bottom line, and project plans update automatically when milestones shift, rather than waiting for a project manager to manually reconcile the data. New team members get a full briefing the moment they’re assigned — not because someone wrote it, but because the system has already listened to the client calls, absorbed the scope, and understands the history.
Even something as mundane as timesheet compliance becomes a different problem. Instead of chasing people to log their hours, the system compares what’s in their calendar against what they were scheduled to do and helps them fill in the record accurately, without ever opening a separate application.
The thread running through all of it is knowledge that used to be locked away.
“I’ve relied on that project manager that’s got the experience of 20 years, that knows the customer and can spot there’s a risk,” shares Sarah. “We can’t afford to rely on that anymore. We have to amplify that talent and their expertise.” The Expertise Agent is what makes that amplification possible at scale, by using and contextualizing what your best people know and leveraging it on every engagement.
“The real moat stops being headcount. It’s now how do we grow revenue and deliver higher value services and deliver outcomes to customers more quickly without that linear headcount growth? The way that you achieve that is by making sure that your firm’s knowledge or expertise lives in a system that services that to everyone — it’s not just in a few people’s heads.”
– Sarah Edwards, Chief Product Strategy Officer
The Expertise Era
For years, professional services firms have competed on efficiency to drive utilization, protect margins, and optimize the billable hour. And that work isn’t going away. But it’s no longer enough to differentiate.
What sets firms apart isn’t time spent on a piece of work. It’s the knowledge and expertise people bring, and how effectively that gets shared across the business.
This is the promise of the Expertise Agent. It’s not about achieving faster task execution, but about developing a smarter operation, where every project benefits from what the business has already learned, every team member has the context they need from day one, and unpredictable projects stop being the cost of doing business.
With one, powerful tool at your fingertips, you’ll always deliver amazing.