StruckBox Promote
AI-judged promotional assessment centers
Run a defensible promotional cycle in two weeks for a fraction of an outside vendor's cost. Identity-blind AI scoring against NFPA-anchored rubrics. Hash-chained audit trail for every cycle.
vs. the alternatives
| Promote | FPSI / Wollack / M&M | Chief's gut |
|---|
| Per-cycle cost | $750-$2,500 | $3K-$25K | $0 (lawsuit risk) |
| Setup time | 2 weeks | 8-12 weeks | 30 minutes |
| Scoring consistency | Deterministic (ICC ≈ 1.0) | Human ICC ≈ 0.70 | Politics |
| Identity-blind | Built in | Optional | No |
| Audit trail | Hash-chained ZIP | PDF + Excel | None |
| Adverse-impact reports | Auto on every cycle | Extra fee | None |
| NFPA 1021 / 1002 anchored | Yes | Yes | No |
How it works
Step 1
Sign the job analysis
20-question wizard pulls NFPA 1021 (or 1002 for D/E) JPRs as a starting point. You weight them for your dept and sign. PDF generated, content hashed.
Step 2
Set up the cycle
Pick rank, dates, candidates from your roster, and exercises. Sign 5 attestations. Cycle locks; rubric is now immutable.
Step 3
Candidates take it
Knowledge test, oral interview, in-basket, tactical IC sim, role-play, writing, presentation — by rank. Voice-driven where appropriate. Identity-blind tokens (Candidate-A, B, C) — the AI never sees a name.
Step 4
Publish the list
Multi-AI panel scores. Disagreements flag for human review. Eligibility list with bands and confidence intervals. Adverse impact computed. Download the audit ZIP for legal.
Pricing
Per-cycle, billed when you actually run a real promotional process. Practice mode is included free with your StruckBox dept tier.
Driver/Engineer
$750
Knowledge test only
Lieutenant
$1,500
5-exercise battery
Captain
$2,000
5-exercise battery + multi-company
Battalion Chief
$2,500
6-exercise battery + multi-incident
Founding-customer waiver available for the first 5 depts. Email
brian@firstdueco.com.
FAQ
Is this defensible against a Title VII challenge?
Promote is more defensible than what most small depts do today. Hash-chained audit log, identity-blind scoring, NFPA-anchored rubrics, and auto adverse-impact reports per cycle. Your dept still owns local validation responsibility — we're a tool, not a validation service.
Does the AI know who the candidates are?
No. Each candidate is given a per-cycle deterministic token (Candidate-A, B, C). The scoring panel never sees names, emails, or photos. Runtime fingerprint sweep blocks any leak before it reaches the model.
What if the AI gets it wrong?
Three independent rater profiles (tactical, leadership, policy) score each attempt in parallel. If they disagree past threshold on any dimension, the attempt flags for human review. Candidates can appeal scores within 14 days; chief or external IO reviewer can adjust.
What about NFPA copyright?
We cite JPR identifiers (e.g. "FO-I 4.6.2") and use paraphrased duty-area summaries written in-house. We never reproduce verbatim NFPA text in candidate-facing UI or rubric anchors.
What about my dept's SOPs?
Upload them as PDFs. We parse, chunk, and embed them. The policy rater references your specific SOPs when scoring — not generic ones.
Ready when you are
We're onboarding the first 5 depts now. The waiver covers your first cycle in exchange for a short case study and a follow-up call.
Talk to Brian