Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for private equity associates

Customer Cohort Diligence Agent

Verified retention, cohort and concentration figures that either support or contradict the company's model

AnalysisFinance
See how it worksGet access
Customer Cohort Diligence Agent: what goes in, what the agent does and what you get

What it does

A company claims 95% retention, but customer data is messy: duplicate accounts, merged customers, renamed entities and one-off projects mixed with subscriptions. Averages hide real churn. This agent reads the customer and revenue file and cleans it, merging likely duplicates and tagging revenue types. It builds cohorts by start period and tracks how much revenue each cohort keeps over time, then measures retention, expansion and concentration in the top customers. It compares the results with the retention shown in the company's model. When something does not agree, it lists the inconsistent rows and asks the company for corrected data, then recomputes with the new file. The associate approves what goes into the findings memo. Edge case: a customer that split into three accounts looks like two churns and one new logo, and the agent merges them.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Customer revenue file is received 2 USES A TOOL Load the file and the company model 3 DOES Merge likely duplicate customers and tag recurringversus one-off revenue 4 CHECKS THE RESULT Do the totals tie to reported revenue within 1%? If not: find the rows that do not tie and requestcorrected data. Back to step 2. 5 DOES Build cohorts by start period 6 DOES Compute gross and net retention and concentration ofthe top 10 customers 7 DOES Compare with the retention in the company's model 8 CHECKS THE RESULT Is the difference under 3 points? If not: list inconsistent rows, request data fixes andrecompute. Back to step 2. 9 DOES Write findings with charts and open data requests 10 YOU APPROVE Associate approves the findings memo 11 RESULT Cohort analysis delivered
Read the steps as a list
  1. Customer revenue file is received
  2. Load the file and the company model
  3. Merge likely duplicate customers and tag recurring versus one-off revenue
  4. Do the totals tie to reported revenue within 1%?If not: find the rows that do not tie and request corrected data. Back to step 2.
  5. Build cohorts by start period
  6. Compute gross and net retention and concentration of the top 10 customers
  7. Compare with the retention in the company's model
  8. Is the difference under 3 points?If not: list inconsistent rows, request data fixes and recompute. Back to step 2.
  9. Write findings with charts and open data requests
  10. Associate approves the findings memoThe agent waits here for your OK.
  11. Cohort analysis delivered

How it decides

It trusts cohort results only after duplicates are merged and revenue types are tagged, and it flags any gap with the company's reported retention above a set size.

  • Merge accounts with matching tax ids or addresses
  • Treat projects and one-off fees as non-recurring
  • Flag any customer above 10% of revenue
  • Flag retention differences over 3 points

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Merge rules for duplicates
  • Concentration threshold (default 10%)
  • Retention gap tolerance (default 3 points)
  • Cohort period (default quarterly)

What keeps you in control

It always asks you first

  • Findings memo before it goes to the deal team

Hard limits

  • Never changes the company's file
  • Keeps customer names inside the deal team
  • States all merge rules used

It stops when

  • Done: cohort results are reconciled and approved
  • Stop: the company cannot provide a usable file

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA target reported 94% net revenue retention. After merging 38 duplicate accounts and removing $1.1 million of one-off fees, the agent computed 86%. The 8 point gap failed the check. It listed 12 inconsistent rows and requested corrected data. The new file raised the figure to 88%, still 6 points below the model. The associate approved a memo with the gap highlighted.

More agents for private equity associates

Document handling · Management

Acquisition Diligence Request Agent

A current map of which diligence requests are answered, partly answered or contradicted.

7 steps1 check7 jobs
Quality control · Finance

Financial Model Formula Repair Agent

A model that passes all test cases after the repair.

7 steps1 check10 jobs
Reporting · Finance

Founder Reference Check Agent

A comparable, themed reference summary with at least a set number of independent sources per theme

11 steps2 checks2 jobs
Analysis · Reporting

M&A Integration Finance Tracker Agent

Show verified synergy delivery against plan each month and drive action on shortfalls.

11 steps2 checks5 jobs