Industry Guide: Private Equity
How to Use Claude in Private Equity: What Deal Teams Actually Do With It
Less time formatting the deal, more time judging it.
This guide shows private equity deal teams how to work with Anthropic's Claude across the deal cycle: sourcing, CIM analysis, due diligence, LBO modeling, IC memos, and portfolio monitoring. It names the financial services tools Anthropic has actually shipped, cites only published customer stories, and gives you prompts you can paste into your next deal today.
Knowing how to use Claude in private equity comes down to a handful of repeatable jobs: screening inbound CIMs and teasers, turning a data room into a structured diligence tracker, auditing LBO and returns models for broken formulas and hardcodes, drafting IC memos from your own notes, and writing variance commentary on portfolio company KPIs. Claude does the reading, structuring, and first drafts. The deal team keeps the judgment, the numbers sign-off, and responsibility for anything shown to the investment committee or LPs.
Published
Key Takeaways
- The highest-value private equity uses are CIM screening, diligence request lists, LBO model review, IC memo drafts, and portfolio company reporting.
- Anthropic has published financial services connectors, Agent Skills, and agent templates that cover comps, DCF models, data room packs, and teasers.
- Claude reads and structures a data room fast, but every number that reaches an IC memo still gets tied back to the source document.
- Confidential deal data belongs only in a Claude workspace your firm has approved for it.
- Specific prompts win: name the sector, the fund's thesis, the check size logic, and the exact output format you want.
Copy and Paste
Prompts You Can Use Today
First-pass CIM screen
You are a senior associate at a lower middle market private equity fund. Our thesis: founder-owned B2B services businesses with recurring revenue, low customer concentration, and a clear buy-and-build path. Read the attached CIM and give me: a one-paragraph business summary, the five facts that matter most for our thesis, every claim that needs diligence verification, customer and supplier concentration as disclosed, and a pass, maybe, or pursue recommendation with your reasoning. Quote the page number for each fact.
Swap in your own fund thesis. Asking for page numbers makes every fact checkable against the CIM.
Diligence request list by workstream
Draft a due diligence request list for the acquisition of the company described below. Organize it by workstream: financial and quality of earnings, commercial, legal, tax, IT and cybersecurity, HR, and operations. For each request, state why it matters to the investment case and mark it as a first-round or confirmatory request. Flag the requests most likely to surface a valuation issue. Company summary: [paste teaser or CIM summary].
Paste the teaser or CIM summary so the list reflects this business, not a generic checklist.
LBO model audit
Review the attached LBO model as if you were the vice president checking an associate's work before it goes to the deal partner. Trace the sources and uses, the debt schedule, and the returns calculation. List every hardcoded number inside a formula, every circular reference, any case where the balance sheet does not balance, and any assumption that looks inconsistent with the CIM. Do not change the model. Give me a numbered list of issues with the cell reference for each.
Asking Claude not to change the model keeps the review separate from the fix, so you decide what gets corrected.
IC memo first draft
Using only the diligence notes and model outputs pasted below, draft an investment committee memo with these sections: transaction overview, investment thesis, key risks and mitigants, valuation and returns, value creation plan, and open diligence items. Write for partners who will read it in ten minutes. Do not add any fact, number, or comparable that is not in my notes. Where the notes are thin, write [NEEDS INPUT] instead of filling the gap. Notes: [paste].
The [NEEDS INPUT] instruction stops the draft from papering over gaps the committee will find anyway.
Portfolio company KPI commentary
Here is the monthly KPI pack for one of our portfolio companies, with budget, prior year, and actuals for revenue, gross margin, EBITDA, cash conversion, and headcount. Write board-ready variance commentary: for each metric explain the variance to budget in plain terms, separate one-off items from trend changes, and list three questions the operating partner should ask management before the board meeting. Keep it under four hundred words. Data: [paste table].
Include prior year as well as budget so Claude can separate a timing issue from a real trend.
From Anthropic
What Anthropic offers for Private Equity
Claude for Financial Services
Anthropic's financial services offering, announced in July 2025, pairs Claude models with Claude Code and enterprise access, pre-built connectors to data providers including PitchBook, FactSet, S&P Global, and Daloopa, and implementation partners. Anthropic says customer data is not used for training by default.
· Source
Advancing Claude for Financial Services
An October 2025 update adding Agent Skills for comparable company analysis, discounted cash flow models, due diligence data packs built from data room documents, and company teasers, plus new connectors including Chronograph for portfolio monitoring and Egnyte for data rooms.
· Source
Agents for financial services
Ten agent templates released in May 2026, including a pitch builder, model builder, valuation reviewer, and market researcher, each available as a plugin in Claude Cowork and Claude Code. Anthropic also made its Excel, PowerPoint, and Word add-ins generally available.
· Source
In Practice
How Private Equity Teams Use Claude
According to Anthropic's published customer story
Carlyle
According to Anthropic's May 2026 announcement of its financial services agents, Carlyle's Chief Digital Officer described Claude as a key part of the firm's AI technology stack.
According to Anthropic's published customer story
British Columbia Investment Management Corporation (BCI)
In Anthropic's October 2025 financial services update, a senior principal on BCI's private equity team said Claude had sped up how quickly the team gets up to speed on investments.
According to Anthropic's published customer story
Hg
Anthropic's May 2026 financial services post quotes the head of Hg Catalyst describing Claude as a powerful tool for automating complex analysis.
According to Anthropic's published customer story
Chronograph
According to Anthropic's customer story, Chronograph, which builds private equity portfolio monitoring software, reached full adoption of Claude across its staff with high daily usage, and new hires became productive within weeks instead of months.
Illustrative walkthrough
Screening an inbound CIM before the partner meeting
- An associate uploads a newly received CIM and pastes the fund's thesis in a few sentences.
- Claude returns a business summary, the facts that matter for the thesis, and concentration risks, each tied to a page of the CIM.
- The associate checks every page reference and corrects one misread margin figure from a scanned table.
- Claude drafts a short list of questions for the banker covering the gaps the CIM leaves open.
- The associate brings a one-page screen and the banker questions to the partner meeting, and the partner decides whether to proceed.
Illustrative walkthrough
From diligence notes to an IC memo draft
- The deal team collects workstream notes, the quality of earnings summary, and model outputs in one project.
- Claude drafts the IC memo in the fund's standard sections and marks every unsupported point as needing input.
- The vice president fills the gaps, rewrites the thesis section in the team's own words, and ties each figure to the model.
- Claude runs a final consistency pass, listing any figure that appears differently in two places.
- The partner reviews the memo and owns the recommendation to the committee.
Private Equity and Claude: the Published Numbers
| Figure | Context | Source |
|---|---|---|
| 100% adoption across 150+ employees | Chronograph, a private equity software company, per Anthropic's customer story | Source |
| 80% daily active usage | Chronograph's reported daily usage of Claude, per Anthropic's customer story | Source |
| new hires productive by week 2 vs 2+ months previously | Chronograph's reported onboarding change, per Anthropic's customer story | Source |
Core Workflows
Where Claude Fits in the Private Equity Deal Cycle
These are the Claude private equity use cases where deal teams get the most time back, in the order a deal moves.
Deal sourcing and screening
AI in private equity deal sourcing is mostly about triage. Claude summarizes teasers and inbound CIMs against your thesis so associates spend their reading time on the few deals that fit.
CIM analysis
CIM analysis with AI means pulling the business model, customer concentration, margin history, and management claims out of a long document, with page references, so the team can challenge them.
Due diligence
AI for private equity due diligence covers request lists by workstream, data room indexing, contract term extraction, and a running tracker of what is answered, open, or contradicted.
LBO and returns work
AI for LBO modeling is strongest as a reviewer: tracing formulas, spotting hardcodes, and explaining sensitivity tables. Your model remains the source of truth for IRR and MOIC.
IC memo drafting
Claude turns diligence notes and model outputs into a structured IC memo draft, so the team spends its time on the argument and the risks rather than on formatting.
Portfolio monitoring
For portfolio companies, Claude drafts KPI variance commentary, board pack summaries, and 100-day plan trackers from the data management already sends you.
The Deal Cycle
How PE Firms Use AI From Teaser to Exit
Ask how PE firms use AI and the honest answer is: unevenly, and mostly on reading and writing. A mid-sized fund sees far more teasers and CIMs than its associates can read closely. Claude reads the full document, summarizes it against your thesis, and lists the claims that need verification. The associate still decides whether the deal deserves a call with the banker, but the decision rests on a structured summary instead of a skim.
Once a deal moves forward, the data room becomes the bottleneck. Anthropic's October 2025 financial services update added an Agent Skill for due diligence data packs that processes data room documents into spreadsheets with financial information, customer lists, and contract terms. Whether or not your firm uses that skill, the pattern is the same: ask Claude to extract specific fields from specific documents into a fixed table, then spot check the extraction against the source files.
Modeling is where teams need the most discipline. Claude can explain an LBO structure, write Excel formulas, build a sensitivity table layout, and audit an existing model for errors. It should not be the only thing standing between a spreadsheet and an investment committee. The strongest pattern is to have Claude review a model the associate built, rather than trusting a model Claude built without review.
After close, the work shifts to the portfolio company. Operating partners use Claude to turn monthly KPI packs into board commentary, to draft value creation plan trackers with EBITDA bridges, and to prepare questions for management. Near exit, the same skills help with vendor diligence preparation and management presentation drafts.
Anthropic's Finance Tooling
Claude Skills for Private Equity and What They Cover
People searching for Claude skills for private equity usually mean the Agent Skills and plugins Anthropic has published for financial services. According to Anthropic, the October 2025 update included skills for comparable company analysis, discounted cash flow models with WACC and sensitivity tables, due diligence data packs, company teasers and profiles, earnings analyses, and initiating coverage reports.
In May 2026 Anthropic released ten agent templates for financial services, each shipped as a plugin in Claude Cowork and Claude Code. The list includes a pitch builder, meeting preparer, earnings reviewer, model builder, market researcher, and valuation reviewer. Anthropic also listed connectors to data providers such as PitchBook, FactSet, S&P Global, Daloopa, Chronograph, and Egnyte, which matter for private equity because they bring market data, portfolio monitoring, and data room content into the same workspace.
None of this replaces your firm's process. Skills and connectors reduce the setup work, but the reviewer at the end is still a person who knows the deal. If your firm has not enabled these tools, every workflow on this page still works in a standard Claude chat with documents you are permitted to upload.
Choosing Tools
Claude vs ChatGPT for Private Equity, and the Best AI Tools for Private Equity
The Claude vs ChatGPT for private equity question matters less than it looks. Both are capable general assistants, and both change every few months. The fair test is your own work: take one CIM you have already screened, run the same prompt in each tool, and compare accuracy, page references, and how each handles what the document does not say.
Lists of the best AI tools for private equity mix three different things: general assistants such as Claude, deal-specific platforms built on top of models, and data providers that connect to them. Decide which job you are solving first. Reading and drafting is a general assistant job. Deal pipeline tracking and LP reporting may already live in software your firm owns, and the right move is connecting the assistant to it rather than replacing it.
Whatever you choose, AI for private equity only works inside your firm's governance. The tool your compliance team has approved, with the data terms your LPs expect, beats a slightly better tool nobody is allowed to use on live deals.
Getting Started
A Practical First Month for a Deal Team
Week one: one CIM
Take a CIM you already screened by hand. Run the first-pass screen prompt from this page and compare the output to your own notes. Adjust the thesis wording until the output matches how your fund thinks.
Week two: a diligence tracker
On a live or recent deal, have Claude build the request list by workstream and a tracker table. Keep updating the same conversation or project as answers come back.
Week three: model review
Give Claude a finished model and ask for an audit, not a rebuild. Log which issues were real. This tells you how much to trust it on the next model.
Week four: share the prompts
Write down the prompts that worked, with your fund's thesis built in, and share them with the team. A shared prompt library is how one associate's habit becomes a firm workflow.
Limits
What Claude Will Not Do for a Deal Team
Claude can be confidently wrong. It can misread a table in a scanned CIM, attach a number to the wrong year, or fill a gap with a plausible assumption. Anything that reaches an IC memo, an LP letter, or a lender needs to be tied back to the source document by a person. Asking for page references and [NEEDS INPUT] markers makes that check fast.
Claude does not know your market the way your deal team does. It has no view on which banker is running a tight process, which management team oversold last time, or how a sponsor in your sector behaves at auction. That judgment is the job, and it stays with the partners.
Confidentiality is the other hard line. CIMs arrive under NDA and data rooms carry customer and employee information. Use only the Claude workspace your firm has approved for deal data, check the data retention and training terms your firm agreed to, and when in doubt, anonymize the company and the figures before pasting.
Practice These Workflows Before Your Next Deal
Reading about how to prompt Claude and doing it well on a live CIM are different skills. The free Nightschool AI curriculum is hands-on from the first lesson, so you build the habits before the deadline arrives.
Frequently Asked Questions
How do private equity firms use AI?
Private equity firms use AI mainly for reading and writing work: screening teasers and CIMs against the fund thesis, building diligence request lists, extracting data from data rooms, reviewing models, drafting IC memos, and writing portfolio company commentary. Anthropic's published customer quotes include Carlyle describing Claude as a key part of its AI technology stack. The investment decision and the numbers sign-off stay with the deal team.
How does AI help with due diligence in PE?
AI helps by organizing the volume. Claude can draft a request list by workstream, index a data room, pull contract terms and customer lists into a table, and track which questions are answered or contradicted. Anthropic has published an Agent Skill that processes data room documents into spreadsheets. Every extracted figure still needs a spot check against the source before it drives valuation.
What AI tools do PE firms use?
PE firms typically combine a general assistant such as Claude with data providers and deal software. Anthropic lists connectors for PitchBook, FactSet, S&P Global, Daloopa, Chronograph, and Egnyte, among others, plus agent templates for pitch building, model building, and valuation review. Which of these your firm can use depends on what compliance has approved.
Can AI build an LBO model?
Claude can lay out an LBO structure, write the formulas, and populate a model in Excel, and Anthropic has published model builder tooling for this. The safer use is review: have Claude audit a model an associate built for hardcodes, broken links, and inconsistent assumptions. Treat any model Claude builds as a draft that a person checks line by line before it informs a bid.
How to analyze a CIM with AI?
Upload the CIM, state your fund's thesis in two or three sentences, and ask for a business summary, the facts that matter for the thesis, concentration risks, claims that need verification, and a recommendation, each with a page reference. Then check the page references. The prompt on this page is a working starting point you can adapt to your fund.
Can AI replace private equity analysts?
It replaces some analyst tasks, mostly first-pass reading, formatting, and drafting. It does not replace the judgment about which deals to pursue, the relationships with bankers and management, or accountability to the investment committee. In practice analysts who use Claude well spend more of their time on analysis and less on production.
Is ChatGPT or Claude better for private equity?
Both are capable, and the answer depends on your documents and your firm's approved tools. Run the same CIM screen or model audit in each and compare accuracy, citations, and how each handles missing information. Also compare data terms and connectors to the providers your firm already pays for, since those often decide the question in practice.
Is AI safe for confidential deal data in private equity?
It can be, inside the right workspace. Use only the Claude plan your firm has approved for deal data and check its retention and training terms. Anthropic states that its financial services offering does not use data for training by default. Outside an approved workspace, anonymize company names and figures before pasting anything covered by an NDA.
Keep going
Related Guides and Courses
Claude in Private Equity, by role
Build the Habit Before the Next Process
Hands-on lessons, real prompts, work you can use on Monday.