AI Case Coaches: The Good, The Bad, And The Solutions

Using AI for MBB case practice can be detrimental. Former McKinsey engagement manager and AI case coach builder Chang Liu tells us when AI is hurting candidates – and how they can prep better for cases.


Editor’s note: This summary was compiled by AI. 

Somewhere tonight, a first-year MBA is running her 40th case practice. She’s talking to a laptop, and the laptop keeps telling her she’s doing great. Today, we explore whether that’s exactly the problem.

THE RISE OF AI IN CONSULTING RECRUITMENT

In recent years, AI has become a significant player in the consulting recruitment process. As Chang Liu, founder of The Case Room and former McKinsey engagement manager, explains, AI tools are being integrated at every stage—from resume screening to interview coaching. Chang notes, “Where is AI not present these days?” This question highlights the pervasive role of AI in how candidates prepare for consulting roles.

With firms like McKinsey and BCG introducing AI-driven tools, candidates are also turning to technology to refine their interview skills. But as those tools become more common, what are they getting wrong?

THE CHALLENGES WITH AI INTERVIEW COACHING

Chang raises some important points about the shortcomings of current AI coaching tools. While they can assist with polishing resumes and practicing quantitative skills, they often fall short in delivering the nuanced feedback necessary for behavioral interviews and case prep.

“When you see the company you want to work for introducing AI in the interview stage, you know that this is where you need to incorporate the technology too.”

However, candidates often emerge from AI training sessions feeling overconfident or misled about their readiness. Chang points out that AI tends to reinforce weak answers with flattery instead of challenging candidates to defend their thought processes: “AI tends to say, ‘that is still a really good starting point,’ which is not the type of rejection you’ll hear in an interview.”

Conversely, AI can be overly harsh on candidates’ math skills, focusing too much on specific answers rather than the reasoning behind them. “In a real interview, it’s not just about getting the right number; it’s about how you got there,” Chang emphasizes.

BUILDING A BETTER AI TOOL

Recognizing these issues, Chang set out to create a more effective AI case coach that mimics real-life interview scenarios. Her process involved extensive testing and refinement, aiming to create an AI that not only challenges candidates but also encourages them to think critically—a crucial skill in consulting interviews.

“I have a lot of friends who continuously reach out to me for me to review their resumes… and I realized students are starting to sound a little bit similar.”

Chang’s insights reveal a clear gap in the market: a need for an AI tool that supports individuality and critical thinking rather than promoting a one-size-fits-all approach.

As she continues to iterate on her product, her commitment to helping candidates showcase their unique perspectives and critical thinking skills remains at the forefront.

FINAL THOUGHTS

The conversation with Chang Liu sheds light on the evolving role of AI in preparing for consulting interviews. While AI can streamline the preparation process, it’s vital for candidates to remain aware of its limitations and continue developing their critical thinking skills. As Chang aptly puts it, “Students need to be wary of these things and counterbalance that somehow.”

For those interested in diving deeper into the intersection of AI and consulting, be sure to check out the full episode of Business Casual.

Stream it on your listening platform of choice: https://www.buzzsprout.com/221123/episodes/20044442

Check out our other episodes: https://poetsandquants.buzzsprout.com

Subscribe to our YouTube channel for more: https://www.youtube.com/@PoetsQuants

© Copyright 2026 Poets&Quants. All rights reserved. This article may not be republished, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Poets&Quants, please submit your request HERE.