Home
Back to Blog
GUIDEBasic

Breaking Into PE or IB: The AI-Native Analyst Toolkit

How to use Claude Code to prepare for private equity and investment banking recruiting without crossing the line: technical drilling with a commit-first rule, paper LBO practice with code-checked answers, a public deal case study, and the rules on AI in tests and interviews.

November 19, 202612 min readClaude Code Playbooks
break into private equityib recruiting preppe interview preparation aipaper lbo practiceinvestment banking interview prepfinance recruiting aiclaude code finance

This post contains affiliate links. If you sign up through one, we may earn a commission at no extra cost to you. How we handle this.

Recruiting for investment banking and private equity rewards one thing above everything else: being fast and right on technical material under pressure. That is a skill you build by practice, and AI is very good at generating practice. It is also very good at helping you cheat yourself, by producing polished answers you can't reproduce in a room with no laptop.

This guide is the toolkit for using Claude as a training partner, not a crutch. The principle is simple: AI helps you prepare; it doesn't help you perform. Everything below is designed to make you better at the part where you are on your own.

Rules of the Road

  • Follow each firm's rules on AI in assessments. Timed modeling tests, case studies and superdays usually forbid outside help, and AI is outside help unless the instructions say otherwise. Read the instructions every time. Being caught is career-ending, and being unprepared without it is survivable.
  • Policies are moving in both directions. The Financial Times reported in September 2026 that UBS wants graduates and interns joining its 2027 global banking and markets intake to show how they would use AI responsibly. We saw that through secondary coverage. Other firms restrict AI tightly. Don't assume, and ask your recruiter if it isn't clear.
  • Never use your employer's or a client's confidential information in practice material, including from a current or past internship. Build on public companies and closed deals.
  • Don't memorize AI-written answers. Interviewers ask follow-ups, and a rehearsed answer you can't extend is easy to spot.

New Claude workflows for finance professionals, every week

One email a week with the newest playbooks. Free, like everything here.

No spam. Unsubscribe anytime.

What Interviews Commonly Test

Recruiting guides describe a similar structure across IB and PE: technical questions on accounting and valuation, a deal walkthrough, a paper LBO in a first round, a timed modeling test (often an LBO with a short write-up) later, and fit questions throughout. Timelines are the unstable part. The process has moved around in recent years, with private equity recruiting restarting after a pause and shifting its schedule, and most of what you will read about dates comes from coaching sites with their own interests. Treat any specific date as an estimate, and check current guidance from your school's career office and recent press before you plan around it.

The Toolkit

1. A technicals examiner with a commit-first rule

The failure mode of AI tutoring is reading the answer and thinking you knew it. Prevent that by making Claude withhold the answer until you commit.

Quiz me on accounting and valuation technicals for an investment banking analyst interview. Ask one question at a time and wait for my answer. After I answer, tell me whether I was right, what I missed, and one follow-up an interviewer would ask. Track the topics I get wrong and come back to them. Don't give me the answer before I have tried.

Keep a list of what you miss. That list is your study plan, and it is more valuable than any generic set of questions. Cross-check anything Claude states as fact against a textbook or a course, because a confidently wrong explanation of a technical point is the most expensive thing a tutor can give you.

2. Paper LBO practice, with answers computed by code

A paper LBO is a mental-math exercise: given an entry price, leverage, growth and an exit multiple, estimate the return. Language models are unreliable at arithmetic, so don't trust them to produce the answer key. Have Claude generate the case and compute the answer in code, and then you check yourself.

Generate a paper LBO case with round numbers: LTM EBITDA, entry multiple, debt as a multiple of EBITDA, interest rate, EBITDA growth, cash conversion and exit multiple. Don't show the answer. Write the solution as Python and run it. Reveal the result only after I give you my MOIC and IRR, then show where my estimate went wrong.

The shortcut every candidate should have memorized comes straight from the arithmetic: with no interim distributions, IRR is the multiple raised to one over the years held, minus one. Over five years, 2.0x is about 15%, 2.5x about 20% and 3.0x about 25%. The tested engine in our LBO walkthrough is a good base for the code Claude writes, and it also shows where the return comes from: EBITDA growth, the multiple, debt paydown and fees. Interviewers like candidates who can say which of those is doing the work.

3. A public deal case study, in your own words

A case study on a real deal is the best proof you can generate before you have a job. Choose a closed take-private, read the proxy statement and the press release, and then build it:

  • Sources and uses and a simple LBO with the LBO Model Builder as your shadow build, after you have built yours.
  • A one-page screening memo of whether the target fit a plausible fund's criteria, using the Deal Screening Memo format.
  • A mock book section, to learn what buyers expect to read. The CIM Builder shows the structure. Draft from public information only, and label it as an exercise.

Then be able to rebuild the core of it from a blank page. If you can't explain the model without Claude, it isn't yours yet, and an interviewer will find that out within two questions.

4. Behavioral stories you can actually tell

Fit questions decide more offers than candidates expect. The Interview Mastery System analyzes a job description against your resume, predicts likely questions and drafts STAR-method answers from your real experience. Use it to structure your stories, then say them aloud until they are yours. A STAR answer should sound like you, so strip anything you wouldn't say.

5. Market awareness and a foundation

Interviewers ask what is happening in markets and whether you have a view. A free Koyfin plan is a reasonable place to follow charts, sectors and macro series without paying for a terminal. And if your accounting and modeling foundation is the weak link, a structured course such as CFI can give you a path. We cover who should skip it in Is the CFI All-Access bundle worth it? Free routes (company filings, textbooks, the playbooks here) cover the same ground if you have the discipline.

A Six-Week Prep Plan

WeekFocusOutput
1Accounting and valuation technicalsA list of the topics you miss
2Three-statement and DCF by handA balancing model and a DCF
3LBO mechanics and paper LBOs20 timed paper LBOs, checked by code
4Public deal case studyModel, memo and a deal walkthrough
5Behavioral stories and the "why" answersSix rehearsed stories and your pitch
6Mocks with a personTwo live mock interviews with feedback

If you have more time, the 90-day curriculum builds the foundation these six weeks assume. For the tool landscape you will meet on the job, see AI tools for investment banking analysts, and for the bigger career question, our take on whether AI will replace the analyst.

One last point, because it is the one that matters on the day. A person across the table is testing whether you can think, and what they remember is how you handle a question you didn't expect. Practice with AI as hard as you like, and make sure the person who shows up is you.

This article is for education and isn't career advice. Recruiting processes, timelines and firm policies on AI change often and differ by firm and region, so confirm them with the firm, your school's career office and recent sources. We may earn commissions from some products mentioned.