Working on consulting projects has long involved creating slides — lots and lots of slides.
At McKinsey, AI has allowed consultants to pare back their reliance on PowerPoint.
Kate Smaje, McKinsey's Global Leader for technology and AI, told Business Insider that she's seen usage of PowerPoint drop massively within a couple of months as employees have begun vibe-coding with AI tools — both in the number of presentations they're creating and time spent using the program.
Beyond client presentations, consultants often use PowerPoint as a project-management tool: a working deck that serves as a running compendium of recent research and next steps, and is sent to clients at each week's close.
One McKinsey consultant has created a new approach: an AI-assisted website that acts as a central project hub for clients and the McKinsey team.
Louis-Charles Généreux, an engagement manager at McKinsey, built the website, which he calls the "client visualization hub," for his current project with a North American cable company.
The project involves about 70 people who need to stay informed in real time. Previously, Généreux would have managed that through the slide deck approach. But it wasn't without its problems, he said.
Once a deck was sent out, version control issues began. As updates were added and shared in email threads, multiple versions might circulate, leaving people working from different versions of the deck with varying understandings of the work.
Nor was scrolling through hundreds of PowerPoint slides the most cohesive way to find information.
"The traditional issue that I faced a lot, especially as an earlier engagement manager, was me being in a meeting and referring to it, and others having missed that," said Généreux.
The AI website fixes that by making the latest work more searchable, structured, and available in one place.
It has saved the team time and reduced the disconnect between engineering teams, product owners, and senior executives, said Généreux. "Everyone, irrespective of their knowledge or skills, sees the exact same thing."
Creating the website
To create the website, Généreux used Platform McKinsey, the consulting firm's internal self-service store for approved products. He found a deployment product that could host the site securely and keep it live.
The team had a repository of "dozens and dozens" of HTML files, including analysis, visuals, tables, and text that they used AI to help package into interactive web pages for the site.
It was then launched through a firm-approved system using McKinsey credentials on Cloudflare, which allowed access to be limited to those involved in the project.
AI also helps keep the website up to date. As the project evolves, the site is updated in real time, and at the end of the week, the system generates a podcast-style summary and a memo for people to consume.
PowerPoint has not disappeared, said Généreux, but it is now more of a final output — for a presentation, email, or memo — rather than the place consultants sign in for the day to work.
The real work, he said, is increasingly happening in AI tools — where consultants perform analysis, get challenged on their thinking, and then turn the work into whatever format is needed.
AI is the 'lifeblood' of consulting work
The consulting industry is at the center of the AI shift sweeping through workplaces. Leading firms like McKinsey, BCG, Deloitte, and IBM are advising companies on how to adopt AI while using the technology to overhaul their own work.
The shift is creating questions around consulting pricing models, talent, training, and where value is created — particularly as agents take on knowledge work once handled by humans.
Earlier this year, McKinsey's CEO Bob Sternfels said the firm's headcount comprises 40,000 humans and 25,000 AI agents. Five months later, Smaje told Business Insider that the number of agents is already "many multiples of that," and AI is becoming part of the "lifeblood" of how McKinsey works.
The shift isn't entirely new, she said, explaining that consultants' value has been more than simply producing research for a long time, and the profession has been changing since before the AI wave hit. Their value lies in judgment, pattern recognition, conceptual problem-solving, and helping clients act on their answers, she said.
What AI is now doing, Smaje continued, is shining a light on where and how value is really created for clients, and speeding up the whole process.
At McKinsey, AI has accelerated the early problem-solving cycle, turning what used to be a team's "week one answer" — the first hypothesis after a week of research and analysis — into an "hour one answer."
That frees up consultants to spend more time testing deeper parts of the issue tree rather than waiting days to get oriented, Smaje said.
In practice, Généreux said his team was using the added time by holding "dojo sessions or hackathons" a few times a week to explore ideas that back in the day would have been remote hypotheses.
"We now have the time to do that and the tools to very quickly determine whether or not that's worth exploring," he said.
