WIP 👷🏗️
Over the past ~5 years I’ve been involved with MCC (Minerva Consulting Club). First as co-founder and co-chair, and now as an external advisor. I’ve done and given more than 100 case studies to prepare myself and others for jobs and internships at management consulting firms.
I am currently working on an AI-powered chatbot to help consulting applicants prepare for interviews in a more flexible manner. The chatbot is built on bubble.io. My vision is to have multiple case studies that users could practice on, get fully personalized feedback, integrate graphs/figures, and use voice and audio to simulate a conversation instead of a chat.
For the time being, I have shared a demo version linked below and screenshots of some (very early) results and prototyping:
Initial system prompt (non-user facing) for the chatbot to have context, know its role as an interviewer, and have the necessary quantitative information for the case study
E.g. Market size, population, % of permit holders, etc.
Start of case interview. Chatbot provides initial prompt, and answers general context question for the user. Gives small amount of feedback based on user’s recommended approach to case.
Continuation of case study. Chatbot gives feedack on certain assumptions, pushing back on unreasonable ones.
Continuation of case study. Chatbot gives feedack on certain assumptions, pushing back on unreasonable ones.
[skipping to the end of the case…]
Chatbot provides succint feedback to the candidate - pointing out math mistakes as well as areas to improve on despite no obvious errors.