Private Markets AI Gtm — AI skill for Claude Code

Account-Based GTM for AI Agents in Private Markets

A go-to-market engine for a company selling AI agents to US private equity firms. An AI research agent studies each firm on the web, scores the account on real, sourced buying signals, picks the right people by role, and writes outreach that pitches each person the agent that solves *their* problem.

**Case study.** "Acme Agents" is a placeholder product; set your own in `config/icp.yaml`. The sample accounts are fictional.

The problem

Selling to private equity is different from selling to SMBs:

• The market is small and high-value. There are a few thousand US PE firms, and one closed deal can be a six-figure contract. Mass email burns the market; every account needs research.
• One firm, many buyers. A Managing Partner cares about deal flow, an Associate about the hours from CIM to IC memo, an Operating Partner about the portfolio, a Head of Data about security. One generic email fits none.
• Timing matters. A firm that just closed a fund, made a platform acquisition or hired a Head of Data is far more likely to buy now.

Doing this research well by hand can easily take an SDR most of an hour per firm.

The approach

flowchart LR
    A[Apollo
US PE firms] --> B[Research agent
Claude + web search]
    B --> C[Account score
fit + sourced signals]
    C --> D[Personas
3 people per firm]
    D --> E[Role-specific
messages]
    E --> F[Brief + CSV
for review]
    F -->|--live| G[Saleshandy
4-step sequence]

1. Account sourcing. Apollo organization search for US private equity firms with 11-500 employees.
2. Research agent. Claude Opus 5.5 uses web search and web fetch to read the firm's site and recent news, then produces a structured brief: strategy, sectors, check size, latest fund, and signals from the last 18 months (fund closes, platform deals, add-ons, exits, key hires, AI initiatives). **Every signal must cite