01 / The starting point
The founder saw the opportunity. The team needed a way to use it.
Quiet Light helps founders sell online businesses. When its founder came to Fleet of Geniuses in 2025, the initial request was AI training. He was already experimenting himself; the challenge was helping a team with different levels of experience apply AI to real deal work.
Buyer outreach offered a useful place to begin. Preparing a campaign meant interpreting a deal book, finding plausible acquirers, researching their interests, locating contacts, and drafting individual messages. The work depended on context and judgment, with substantial manual preparation around both.
We began alongside the person doing that work, building around her process and using actual campaigns to refine the system.
02 / What we built
A connected path from deal book to reviewed outreach.
The outreach engine takes a deal book, researches potential acquirers on the web, evaluates their fit against an evidence-based rubric, enriches decision-maker contacts, and prepares personalized messages. The operator reviews the drafts and sends them.
In 2026 we rebuilt the pipeline as Quiet Light Unified, using Claude Code and the Claude Agent SDK on a Quiet Light-paid server. Quality checks and a visible workflow help the operator guide campaigns and identify work that needs attention.
A related buy-side database gives the business-development team a growing research resource. Research loops examine filings and acquirer websites to extract acquisition mandates: what a buyer is looking for, where it invests, and the kinds of businesses it will consider. The application is deployed in Quiet Light’s own hosting behind company sign-in.
Research and fit
Find candidates and evaluate them against the deal, with source evidence behind the assessment.
Preparation and review
Locate contacts and prepare messages for the outreach lead to judge, edit, and send.
Reusable buyer knowledge
Keep acquisition mandates available for future listing-specific research.
03 / What changed
The operator can run the work herself.
By her fifth campaign on the original pipeline, the buyer-outreach lead was running it independently. That operating role continued through the 2026 rebuild. The dashboard now shows substantially more monthly outreach volume handled by the same one-person function.
Campaign research and drafting now follow a repeatable workflow, while targeting decisions, message review, and follow-through stay with the team. The operator can direct more work through the system and apply her judgment where it matters.
04 / The team’s capability
Advisors are becoming builders, too.
As the engagement developed, Quiet Light put more emphasis on helping advisors build directly with AI. The program took shape in May 2026; weekly office hours and monthly show-and-tell sessions began in July.
Advisors have created deal trackers, offer-comparison tools, client-interview skills, and other aids for their own work. These sessions give people a place to bring a real problem, try an approach, and share something useful with colleagues.
Other FOG builds include a searchable deal-debrief hub and a Slack tool for preparing seller improvement decks. Access and adoption are still developing across those tools; an advisor inbox assistant remains in shadow-mode evaluation. The outreach engine provides the clearest example of sustained daily operation.
Inside the systems
Claude Code + Claude Agent SDK
Quiet Light Unified coordinates research, scoring, contact enrichment, and drafting, with automated quality checks and a human review step before outreach is sent.
Client operation
The outreach lead runs the pipeline on a Quiet Light-paid server. The buy-side application uses company sign-in in Quiet Light’s hosting environment.