We recently helped a global payments technology company move off ServiceNow and onto Jira Service Management, a shift that touched how thousands of their employees did their daily work. A few years ago, mapping every one of those existing processes by hand would have burned through the client's entire budget, and then some, before we ever got to the part that mattered. This time, AI got us through that analysis in a fraction of the time.
That didn't shrink the size of the team on the engagement, and it didn't shrink the engagement itself. It changed what the team spent its time doing. Instead of weeks buried in process documentation, our consultants spent that reclaimed time in the room with the client's own teams, asking sharper, data-backed questions earlier in the process, about what the new workflow actually needed to look like for the people using it every day. The result was a higher-quality solution, not a quicker one, built with a client who trusted it because they had helped shape it the whole way through.
That story is the part most of the "AI is coming for consulting" conversation gets backwards.
The industry is absorbing AI fast, and the entry-level, hours-billed model is shrinking with it. That part is real, and it is not slowing down. But the conclusion a lot of people jump to, that organizations therefore need fewer consulting people, has it backwards. What our engagement needed was not fewer people. It was the same people spending their time differently, on the work AI still cannot do.
Start with the part almost nobody is saying out loud: an AI tool is only as unbiased as the people who configured it, and those people work inside the organization being evaluated. An internal team running its own AI-assisted process review will, without meaning to, point the tool away from its own sharpest edges. Nobody sets up their own AI to indict their own department. That is not a flaw in the tool. It is a structural limit on anything built and run by the system it is assessing. On the payments engagement, we were not just running an AI tool against their data. We were asking our own questions, shaped by having led this exact kind of ServiceNow-to-JSM transformation across a wide range of industries, questions their internal team had little reason to ask because they were inside the system being changed.
Organizations are built to serve people, not machines, and a transformation touching thousands of employees is really a few thousand individual conversations wrapped up in one project plan. The client in this engagement knew what they wanted on paper: a less expensive, more modern service management platform. What kept them up at night was how to get several thousand people through that change without grinding the business to a halt. That is not a problem AI solves by itself. It is a change-management problem, and it gets solved by people who know how to make a room full of anxious employees feel like the change is happening with them, not to them.
Speed to insight and speed to belief are two different things, and AI only closes one of those gaps. AI can compress weeks of process mapping into days. It cannot compress the trust it takes for a frontline team to actually adopt a new workflow instead of quietly working around it. Every hour AI saves on analysis is an hour a consultant can spend earning that trust directly, in the room, instead of behind a laptop. That is exactly what happened on this engagement, and it is why the solution held up after we left.
None of this is an argument for smaller teams. It is an argument for the right mix. When manual process analysis ate most of a budget, firms staffed up to produce it. Now that AI compresses that layer into a fraction of the time and cost it used to take, the real value has shifted to something AI cannot supply on its own: judgment about what the analysis means, and the credibility to walk thousands of employees through a change they did not ask for. A firm that reads this moment as "do more with fewer people" will hollow out the exact thing clients are paying for. We read it the other way. It is what let us put more senior time in front of this client, not less, at the moments that mattered most.
We are not interested in defending a model where junior staff spend weeks on process documentation AI can now produce in days. That part should change, and it is. What does not change is how many of the right people it takes to turn faster analysis into a transformation an organization actually trusts and adopts.
AI gives you speed. It still takes a human to give you sight.
If your team is trying to figure out where AI belongs in a transformation and where it doesn't, we'd love to talk. Praecipio helps organizations pair faster analysis with the judgment and trust it takes to make a change actually stick.
Brian Nye is VP of Strategic Initiatives at Praecipio.
