You don't need a strategy deck. You need someone who has been the client, the operator, and the engineer inside the same business — under live regulatory pressure — and ships working systems with receipts.
Most AI consultants have never had to keep a real business alive while the regulatory ground moved under it.
I have. 25+ years as the client, operator, and engineer inside one regulated business (Texas hemp). 20+ months doing FDE work inside that business while Texas shut down the hemp industry in 2026. The pivot is happening in public, in real time.
Reggie & Dro — licensed Texas cannabis retail, Stone Oak, San Antonio — is the proving ground. This isn't advised from outside. It's built from inside.
| Claim | Status |
|---|---|
| 25+ years running a regulated business | ✓ Public bio — github.com/reggieanddro |
| 20+ months, 100-150+ hr/week AI research sprint, inside own business | ✓ Public bio — GitHub + @aicrisiscoach |
| Licensed Texas cannabis retail operation | ✓ Live — reggieanddro.com |
| Currently navigating live regulatory shock (2026 TX hemp shutdown) | ✓ Named publicly, in real time |
| Local, sovereign voice/TTS pipeline running | ✓ Proven — method + reproduction in toROI_HOT.md |
| "Golden Voice" / DSH full pipeline live | ✗ Not yet proven — do not lead with this until it returns 200 |
| Standalone deployed profile site (jesseniesen.com) | ✓ This page — deployed independently |
This table is the proof architecture: real systems, real credentials, no claim that outruns its receipt. The gap between "GitHub bio says FDE" and "site was not its own deploy" was the first thing an FDE audit would flag — which is a stronger opening line than pretending it was already closed.
One workflow. One constraint. One number that hurts. One deployment that proves it moved.
90 minutes + prework. Client brings one workflow, its owner, and the number that hurts.
Deliverable: workflow/decision-path map, constraint diagnosis, baseline metric, go/no-go recommendation.
Structure: fixed fee, credited toward next stage if they proceed.
"Don't buy a transformation program. First prove which bottleneck is worth attacking."
2 weeks. One workflow, one owner, one primary outcome metric.
Deliverable: current-state map, instrumented baseline, minimum viable system, acceptance test, 30-day deployment map, pass/fail readout.
Done condition: a named operator on the client's team can run the new path, with a receipt proving it worked.
6–8 weeks. One business-critical workflow, shipped to production.
Deliverable: production integration, human-in-the-loop approval design, audit evidence architecture, operating dashboard, runbook, team enablement.
Boundary: no open-ended scope creep — new workflows are a new engagement.
90 days, renewable monthly. Weekly operator review, biweekly executive review.
Deliverable: production optimization, failure analysis, next-workflow prioritization, vendor/architecture challenge function, quarterly roadmap.
Pricing: held back until 3–5 Signal Scans have sold — priced against the cost of the bottleneck (recovered revenue + avoided loss + labor/cycle-time reduction + risk reduction), not against hours coded.
Best-fit buyer: Founder/CEO, COO, or Head of Compliance at a company in a regulated or operationally dense industry — cannabis/hemp and adjacent regulated retail are the credible beachhead, because that's the business this offer is proven inside of, not advised into from outside.