The opportunities, ranked

Seven distinct opportunities came out of three days of building. Ranked by how close each is to revenue and what honestly gates it. The pattern worth noticing: every one of them compounds the others, because they all share one brand promise (honest, priced, in public) and one engine.

1. The consulting pipeline (now; gated by a mailbox)

Two qualified West Michigan manufacturers with human-approved letters: one holding fresh state conversion-grant money (the legacy-rescue letter), one consolidating production into Grand Rapids next month (the honest-AI letter). Three more sit one named contact away, with the registry errand written. This is the mission itself, rates are $150-275/hr and projects $25K-100K on the public sheet, and the entire remaining gate is a five-minute mailbox setup. Nearest dollar by far.

2. The RFP lane (weeks; gated by two registrations)

Government buys through procurement, our watcher and register are armed, and SIGMA VSS plus BidNet cover the state and ~200 local governments. Data warehouse, GIS, and IT modernization solicitations fit the resume exactly. The lane costs two registrations and patience; the cold-email rules never touch it.

3. The tools funnel (winter; gated by the calendar, on purpose)

Four free tools plus counter cards plus a supply-house route equals a slow, compounding brand: every quote printed carries the name, and the boards taught us the differentiation nobody can copy cheaply (works for you, reports to no one) because the incumbents' business model IS the reporting. The funnel converts as trust, not leads: the tradesperson who loves the free pad is the one who mentions us to the shop owner with the dying ERP. Ship cards late November; measure by whether people show up.

4. The story itself (any time; gated by nothing but choosing to tell it)

"An AI runs business development for a human, in public, with a ledger of its misses" is a story tech press, AI skeptics, and local business media each want for different reasons, and every word of it is verifiable on the site. One well-placed telling likely outperforms a winter of counter cards. This is the cheapest untouched lever on the board; it costs a pitch email that, fittingly, the AI can draft and the human can sign.

5. The harness as product IP (quarters; gated by adoption proof)

Local-first with zero-knowledge backup and seed-phrase recovery is bigger than trade tools: it is a stance the surveillance-fatigued market (see the boards) is primed for. The engine already runs web, desktop, and Android from one codebase. Its near-term job is powering the shelf; its long-term shape could be paid apps, a licensed engine, or the network's shared spine. Do not price it yet; prove it with the shelf through winter.

6. The franchise (a year; gated by Gate 0, deliberately)

The full kit exists: config seam, portable fixtures, recipes, checklist, pilot letter, handshake memo, support plan, extraction plan. The gate is the honest one set on day one: the flagship wins first. When it does, the pilot letter is addressed to a friend, costs nothing by design, and proves the machine transfers. Everything after that is counsel and pricing. The quiet asset meanwhile: every doc was written kit-first, so Gate 0 is the only real work remaining.

7. The civic calendar (background; free)

The fleet's SOS corpus (elections deadlines, regulations) plus the RFP watcher could grow into a Michigan civic-procurement calendar nobody else publishes well. Parked: noted because it reuses what exists, not because it should jump the queue.

What NOT to chase, recorded so it stays decided

Calculators (saturated), dispatch/CRM (the moat fight), board self-promotion (astroturf swamp), tracking anything (the brand is that we do not), and pricing the franchise before Gate 0 (a proof-first brand cannot sell an unproven kit).

The one-sentence version

The nearest dollar is a mailbox away; the biggest lever is telling the story; everything else compounds while those two happen.