# Part Beacon — Rollout Plan

## Brand architecture

**Part Beacon** is the parent service: a global parts aggregator and sourcing-intelligence platform. Helicopter/aviation parts remain the first vertical and proof market, but the brand is deliberately broad enough for industrial, marine, electronics, equipment, automotive, and other hard-to-find parts categories.


## Phase 0 — Foundation live

Status: mostly complete.

Deliverables:

- live public site at `https://partsace.acedata.ca/`;
- mobile-friendly layout;
- left navigation;
- inventory-first public view;
- source-safe public inventory JSON/table;
- protected private source-linked hot sheet;
- wanted-part alert form;
- photo/OCR intake path;
- scope/framework/data-model docs;
- daily multi-language supplier discovery cron.

Exit criteria:

- public site is readable on desktop/mobile;
- no source links exposed publicly;
- test wanted/photo intake can be stored privately;
- docs are complete enough to guide build work.

## Phase 1 — Demand capture hardening

Goal: make every user submission useful for sourcing.

Tasks:

- improve field validation;
- add clearer examples for part number/model/component;
- improve OCR extraction and token normalization;
- add spam and duplicate detection;
- generate internal daily wanted-alert digest;
- add a private review page for submitted alerts.

Exit criteria:

- every alert has a stable ID;
- every photo intake has extracted token candidates;
- internal reviewer can see clean records without opening raw JSONL.

## Phase 2 — Supplier/source registry

Goal: keep finding and classifying new sources daily.

Tasks:

- maintain supplier registry with domain, language, country, category, confidence, last seen date, source notes;
- distinguish supplier website vs marketplace vs auction vs broker vs dismantler;
- translate non-English listings to English;
- flag new source domains;
- avoid duplicates;
- preserve source URLs privately only.

Exit criteria:

- daily job can report new suppliers separately from repeat listings;
- English summaries exist for non-English discoveries;
- source credibility notes are available for review.

## Phase 3 — Internal match queue

Goal: connect wanted demand to discovered supply.

Tasks:

- build matching script over wanted alerts, OCR tokens, inventory listings, and supplier registry;
- score matches with reasons;
- output Markdown/JSON/HTML queue;
- mark review statuses;
- generate next-action suggestions.

Exit criteria:

- at least one wanted alert can show match/no-match status;
- every match explains why it was suggested;
- no external contact is triggered automatically.

## Phase 4 — RFQ packet generator

Goal: turn qualified matches into approval-ready supplier packets.

Tasks:

- draft supplier RFQ template;
- draft buyer follow-up template;
- include photos/OCR tokens and documentation requirements;
- include exact recipient/source only in private view;
- keep all sends draft-first.

Exit criteria:

- Simon can approve one exact RFQ packet;
- no send happens without approval.

## Phase 5 — Monetization tests

Goal: validate revenue safely.

Possible tests:

- urgent/AOG concierge sourcing;
- paid saved-search alerts;
- supplier listing placement;
- private broker intelligence reports;
- commission/referral agreements.

Exit criteria:

- one approved monetization test with clear risk boundaries and proof.

## Standing daily rhythm

- 06:30 Pacific: multi-language supplier/source discovery.
- Add or update private review artifacts.
- Translate useful listings to English.
- Update public site only with source-safe summaries.
- Notify Simon only when a real lead/update/blocker exists.

## Approval gates

Explicit approval required before:

- supplier contact;
- buyer alerts;
- public posts;
- payments/commissions;
- exposing source links;
- publishing third-party contact details;
- claiming availability/airworthiness/traceability.

## Phase 6 — Subscription plumbing

Status: started.

Deliverables:

- public plan explanation for weekly free inventory, daily inventory subscription, and part sourcing subscription;
- three-month free launch signup form;
- private subscription signup JSONL store;
- payment preference capture for Canadian dollars / Bitcoin / decide later;
- BTC destination stored privately, with only the selected address ending (`4zsg`) exposed to internal review.

Exit criteria:

- signup API accepts dry-run and real launch-trial records;
- no payment is collected during the launch period;
- full BTC address is not published on public pages;
- daily discovery can later segment free weekly users, daily inventory subscribers, and part sourcing subscribers.