Method
It lives in the decisions you make, in investment committee debates, and in what each PM knows without having to articulate it. An AI tool that does not understand that context will remain generic. This is how we make it explicit.
MAKING YOUR CRITERIA EXPLICIT
Ask a PM what matters in their analysis and you will get a valid, high-level answer, usually close to what the firm already states in its investment philosophy. The more useful criteria emerge when you reconstruct a specific decision. Calibration therefore starts from two sources.
01
Decision history
Your investment decisions over several years, analysed together. They reveal recurring patterns: what tends to precede an investment, what triggers an exit, and which issues repeatedly surface in committee. The patterns are visible. The reasoning behind them often is not.
02
Three to four decisions reconstructed in depth
Recent decisions are revisited with the people who made them. This reveals what the historical record cannot: what was enough to act, what was deliberately discounted, and what could have changed the decision.
A hypothetical case
An industrial equipment company reports its half-year results.
Firm A
Asset-value investor. Sixty positions. Three-year horizon. Buys valuation dislocations and exits as the gap closes.
WHAT EACH FIRM SEES
Firm B
Quality compounder. Eighteen positions. Five-year horizon. Buys durable pricing power and exits when the thesis breaks.
Same results, same source material. What rises to the top is different because each firm has its own hierarchy of what matters.
Your firm has one too. It is simply rarely made explicit.
WHAT THIS BECOMES
The calibration produces a framework that belongs to you. It sets out what matters to your firm, how much it matters, and under what conditions. It contains no company-specific facts: instead, it defines what would be material if it appeared in a company or reporting period. That is what makes it reusable across names and over time.
The tools use that framework directly. Each new publication is read through it, and every conclusion links back to the underlying passage, document and page.
The framework prioritizes attention without making the investment decision for you. It does not hide what it considers secondary either: that information still appears in the output, simply with the appropriate weight.