# missionMATCH Demo Walkthrough

A scripted click path for a live client demonstration. It tells one connected story: the Command Center names a
staffing gap, the platform shows the exact people who close it, screens one of them with a transparent score,
assembles a defensible slate, proves the compliance posture, and lets the presenter ask the AI brain a plain
language question. Every candidate shown is synthetic Demo data.

## Before you start

- The demo runs against the live demo organization, Galapagos Federal Systems.
- On a fresh database, bring the whole demo up with one command: `make demo` (this applies migrations, loads the
  base pool, shapes the three hardest gaps, seeds the vendor scenario, then screens the pool through the frozen
  engine so the intelligence panels are populated). Against the live environment the data is already loaded.
- Log in as the demo owner. Multifactor is a time based one time code.

## The story, screen by screen

### 1. Command Center, the mission picture
Open the Command Center. Point out:
- The readiness banner and the total cleared workforce.
- The named staffing gaps. The three hardest are Intelligence Analyst, Data Scientist, and Cyber Architect.
- Say the line: the platform does not just tell me I have a gap, it shows me exactly who can close it.

### 2. Drill into a gap, WhoCanSupport
Open the Intelligence Analyst requirement (or Data Scientist, or Cyber Architect). The WhoCanSupport panel lists
the ranked candidates who have been screened against it. Each of the three headline gaps has a small set of
proposal ready green fillers at the top, then potential and recruiting required candidates below.
- Point out that the ranking is the frozen engine score, and that each candidate carries a readiness band.

### 3. Candidates, the Win Impact view
Open the Candidates page. The fillers for the open gaps show a HIGH Win Impact, because placing them closes a
named gap on an active opportunity.
- Say the line: this is the connected story, the gap the Command Center named and the people who close it, on one
  screen.

### 4. screenMATCH, the transparent score
Screen one of the green fillers against the requirement. Show:
- The overall score and band.
- The per dimension breakdown, Clearance, Skills, Certifications, Customer experience, Experience, Location.
- The plain language rationale and the clearance intelligence panel.
- Say the line: every number is explainable, and clearance is shown as intelligence, the real check is the
  customer facility security officer at onboarding.

### 5. slateMATCH, the defensible slate
Open the Slate tool and select a requirement with a generated slate. Show:
- The ranked members with their scores and bands.
- The human decision controls, approve or reject. Emphasize the platform never auto rejects.
- The adverse impact readiness readout.
- Say the line: the slate carries its own defense, the score, the rationale, and the four fifths reporting a
  review needs.

### 6. Compliance, the AEDT posture
Open the Compliance page. Show the adverse impact report and the audit posture.
- Say the line: demographics never touch the score, they live in a separate module used only for this reporting.

### 7. Ask the Brain, the differentiator
Open Ask the Brain. Click an example prompt or type one:
- How many candidates hold a TS/SCI clearance?
- What are our hardest staffing gaps right now?
- Who can support the Intelligence Analyst mission?
- Which vendors perform best on our submissions?
Show the grounded answer, the path badge, and the citations under Grounded on.
- Say the line: the brain answers only from our own cleared data, it cites its evidence, and it refuses anything
  it cannot ground or anything about a protected attribute.

## The one sentence close
missionMATCH is the system of record for the question every capture and staffing team asks daily, who can
support this mission, and it answers in a way a compliance officer, a proposal manager, and a contracting officer
can all trust.

## Expected on screen anchors
- Three named gaps: Intelligence Analyst, Data Scientist, Cyber Architect.
- Each headline gap shows a small set of green, proposal ready fillers in WhoCanSupport.
- Those fillers show HIGH Win Impact on the Candidates page.
- The vendor intelligence scenario has four vendors with a spread of submission outcomes.
- The Ask panel returns grounded, cited answers and refuses a protected attribute question.