For decades, consulting firms have made money by packaging advice into slide decks and billing by the hour.
Now, developers are using AI to mimic that process and make McKinsey Copy Consultant available directly in your browser.
Take Vercel’s new “Skills” library as an example. This is an open source repository of approximately 90,000 reusable skills for AI agents. This ranges from copywriting and code reviews to consultant-style problem solving. Vercel is an AI startup valued at over $9 billion that operates a cloud-based AI platform for developers.
A “skill” in the AI sense of the word is a capability that a developer can create or download and give to an AI model or agent to enable it to perform a specific task without having to train the model from scratch.
The idea gained traction after Anthropic introduced “skills” to its chatbot Claude in October. This facilitated the spread of the concept. Since then, developers have started building and sharing skills that can be incorporated into various AI systems.
Business Insider examined Vercel’s skills library and found at least four skills labeled with the term “McKinsey” and 26 skills labeled with the term “consultant.”
The most popular consulting-related skill at Vercel is called “McKinsey Consultant.” It was first uploaded on January 25th and has been installed an average of 445 times per week so far. While this is a significant number, it still doesn’t compare to the most popular agents in Vercel’s library, which have hundreds of thousands of installations.
It also has 200 stars on GitHub, which means it’s quite popular and has passed several security audits. This shows that it is both viable and gaining traction among developers. In other words, users find it convenient.
Vercel’s library describes this skill as a prompting framework (originally created for Claude) that guides the AI through problem definition, hypothesis generation, performing structured analysis, and slide creation, recreating the classic workflow of a typical McKinsey consultant.
Business Insider asked former McKinsey staffers to look into the McKinsey-style agents uploaded to Vercel’s library and see how they stack up against the real thing. Arvind Vasudevan, a former McKinsey engagement manager, told Business Insider that McKinsey lacks key competencies that define a consultant.
“What’s missing is how MBB and strategy consultants add value,” he said in a text message, referring to a group of large consulting firms that includes McKinsey, BCG and Bain. “Most of the value is the questions consultants ask and the conversations that help clarify thinking, uncover unstated assumptions, and ensure deep thinking. Nothing of that is happening with this agent. It’s boilerplate analysis without Socratic questions or thinking.”
They may not be real, but AI agents that mimic the work of consultants are already generating millions of dollars in revenue for companies like PromptQL, an AI enterprise platform launched by open source unicorn Hasura.
The platform helps clients build custom AI analysts by integrating internal data with the underlying models they already use. Once deployed, these AI analysts will be able to perform tasks typically handled by data scientists and engineers, and will be able to continuously learn and adapt to their environments over time.
PromptQL co-founder and CEO Tanmai Gopal told Business Insider that the biggest barrier, or moat, to sales analytics is understanding the relationship between people, data, and revenue.
“The McKinsey team spends weeks inside a company, absorbing how it actually operates, the exceptions, the tribal knowledge, the different definitions across departments. It’s that company-specific context that makes their advice worth millions of dollars,” Gopal told Business Insider.
Gopal said enterprise AI tools often fail because they lack a proper foundation, tend to guess rather than ask questions, learn from feedback, and maintain a common understanding across teams.
According to him, PromptQL aims to address these issues through a shared layer of understanding that adjusts as new input comes in.
“Once a team member modifies the AI, teaches the definition of the AI, or resolves ambiguities, that knowledge becomes permanent and available to everyone. It’s not a semantic layer maintained by data engineers. It’s something that comes out of the conversation,” he said.
The model does not automatically recognize internal nuances such as pricing changes, team-specific terminology, or conflicting definitions of revenue. Gopal added that the real problem is not competency, but a lack of context.
In other words, the consultant’s slide deck wasn’t really the product. That’s their decision, and it’s something the AI is still learning.
Have something to share about how consultants are using AI? Business Insider wants to hear from you. Email Lakshmi Varanasi. lvaranasi@businessinsider.com Or contact her on Signal at lvaranasi.70.
