The Vault of Experts had to start somewhere. When Tom set out to build the first candidate, the choice wasn’t obvious — several expert ideas were sitting in the wings (compensation, finance, blind spot discovery, others) — but one of them had a quietly decisive advantage: it could be tested with no live data.

That expert is the Executive Positioning Expert. Its job is to read how a senior professional is currently positioned relative to a stated career objective, identify the gap between that read and what the objective requires, and produce a structured analysis the advisor uses to shape a Strategic Opportunity Memo. It is not a general career coach. It is not a recruiter. It is not a psychometric tool. It is a specialist that thinks about how this person reads to decision-makers and what would have to shift to move them from current positioning to target positioning. Most of what a senior advisor like Bo already does in their head — but laid out, in writing, with the reasoning made legible.

Why first? Three reasons compounded. The expert sits closest to mAInCharacter’s existing client entry point, so its first output isn’t an experiment in a new domain; it’s an experiment in a domain Bo already has instincts about. The work doesn’t require live financial datasets, market data feeds, or any real-time integrations — a controlled case packet is enough. And it produces something the advisor can use the same day. Three boxes ticked. The other candidates ticked fewer.

The build proceeded in two passes. On May 16th, Tom scaffolded the package — nine files defining what the expert is, what it consumes, what it produces, and how to evaluate it. The master prompt at that point ran on general knowledge alone; the curated source base existed only as a target. Six days later, the second pass: Bo’s Six Sub-Path Client Segmentation Framework — articulated by Bo on April 16th in a strategy session with Gareth — was extracted, approved, and folded into the master prompt as the expert’s first practitioner source. Suddenly the expert wasn’t reasoning from generic positioning theory; it was reasoning from the actual decision tree mAInCharacter runs on.

Think of the current state as a chef’s mise en place. The knives are sharpened. The stations are stocked. The recipes are written. The kitchen is open. What’s missing is the first ticket — a real Client Case Packet against which the expert is actually run, with the output reviewed against the 10-criterion rubric. Until that happens, the expert is technically ready and operationally untested. Confident scaffolding is not the same as confident output.

The next gate is exactly one move: Tom and Bo pick the first case, populate the packet, run the activation, and complete the first review log entry. Likely candidate: a real or composite Strategic Discovery case — the highest-deliberation, highest-propensity-to-pay segment in Bo’s framework, and the most strategically valuable to get right.

Two scaffolded experts as of today where last week there was one — and the new one isn’t a sequel, it’s a sibling.

Earlier in the week, a meeting note proposed rebranding Expert #001 as “Discovery Expert.” Tom considered it. Then he discovered that the better move was to keep Expert #001 exactly where it is and build a brand-new Expert #002 beside it. Renaming would have compressed two different jobs into one and quietly destroyed the diagnostic step downstream. The new arrangement is cleaner, slightly more work, and a more honest acknowledgment that figure out which game this person is playing and figure out how they show up in their market are not, in fact, the same question.

So: the new expert. It does one thing. Hand it an intake packet — a coaching application, a discovery call transcript, a LinkedIn excerpt, the loose collection of material that arrives when someone enters the funnel — and it classifies the person into one of Bo’s six sub-paths (or Camper, for those already pitched in a spot they like; or Unclear, when the signals refuse to converge). Verdict, confidence, supporting signals, disconfirming signals, probes for the next conversation. That’s the whole job. No positioning analysis. No sales qualification. No outreach drafts. All downstream of the verdict — and all dependent on it being right.

Think of Discovery as the triage step. It doesn’t decide what’s wrong with the patient. It decides which doctor the patient needs to see in the first place.

The expert lives a double life. In coaching, it runs first and hands its verdict to Positioning. In the Gareth Sales Operating Loop, it produces pre-call hypotheses and post-call refinements, eventually populating the CRM. The sales loop is the high-volume calibration environment; coaching is the high-leverage application. Same expert, two very different pressures. We’ll see which one shapes its evolution faster.

Tom finished the 9-file scaffold today, mirroring Expert #001’s structure exactly. What is not in place — and is worth saying out loud — is a single real activation. The expert is technically ready and has been tested against, to be specific, zero real intakes. The highest-leverage first case is a stated-vs-actual mismatch: the kind of “I’m just exploring my options” that turns out to be a job search wearing a tie. That’s where Discovery earns its keep, and where we’ll find out whether the master prompt is genuinely useful or merely confidently structured.