Add PITCH.md and WHITEPAPER.md — elevator pitch + external white paper

[tautme]

Why

You asked for an elevator pitch and a white paper to share with interested parties. For an open-source project, the right home is alongside the README — discoverable, diffable, amendable by PR. Two new top-level files:

• PITCH.md (68 lines) — three tiers of elevator pitch (10 sec / 30 sec / 60 sec) plus audience-specific spinoffs for book lovers, bookstore owners, the OSS / civic-tech community, and funders.
• WHITEPAPER.md (320 lines) — a share-with-interested-parties document. Twelve sections, ~2,500 words.

What's in the white paper

§	Section	Purpose
1	The problem	Names the gap, with a competitive-landscape table (AbeBooks, Biblio, Bookogs, Goodreads, OpenLibrary, OSM — why each one falls short)
2	What we're building	The two-layer model: OSM owns places, our DB owns contributions, OSM-id joins them
3	How it works	The contribution loop end-to-end; the on-device OCR call-out as the operating-cost argument
4	Principles	The five KISS / Unix tests, lifted from ROADMAP for visibility
5	Architecture	One-page table; hosting / framework / map / data / backend / OCR / detection / analytics, with the $0–$20/mo cost target
6	Current status	Honest — what's shipped, what's in review, what's documented. Doesn't overclaim.
7	Roadmap	Points at ROADMAP.md as canonical
8	Ethics: bookstores as partners	The contributor-norm "ask, don't take" stance; opt-out; no price scraping; copyright stays with contributor
9	Why open infrastructure fits the problem	Argues the shape: a commercial competitor would silo; a consortium wouldn't form; a national archive would be heavier. KISS + open is the right size.
10	Licensing & contribution	MIT for code; ODbL/CC-BY precedent for data (final terms in progress); how to participate
11	Who we'd like to talk to	Bookstores, civic-tech projects, researchers, funders — without overclaiming partnerships
12	Honest limitations	OCR accuracy, inventory freshness, geographic coverage, contributor base, third-party deps. Surfaced because overclaiming is the worse failure mode for an open project.

Voice & tone

Matches the README: opinionated, plain, no consultant-speak. No "synergies," no vague metrics, no "we are pleased to announce." Limitations and ethical tensions are named explicitly rather than buried.

What the pitch covers

• 10-sec hook: "OpenStreetMap for used and indie bookstores — a free, public map built from phone photos of shelves."
• 30-sec / 60-sec elaborations for when there's room.
• Four audience-specific spinoffs: different framings for a book lover vs. a bookstore owner vs. an OSS contributor vs. a funder. Same project, different doorways.

Accuracy notes

• What's shipped vs in-flight is described honestly — the OCR auto-pick (PR OCR: auto-pick paddle/tesseract by device, with visible UI switch #23) is listed as "in review," not "live," because it isn't merged yet.
• No invented metrics — no "5,000 bookstores" or "1,200 contributors" placeholders. The status section is qualitative; you can edit in real numbers when you want them.
• Partnerships are framed as "who we'd like to talk to," not as existing relationships — Internet Archive, Wikidata, Open Library are mentioned as adjacencies and possible partners, not as affiliations.

Files

• PITCH.md (68 lines, ~400 words)
• WHITEPAPER.md (320 lines, ~2,500 words)

Both unaffected by lint / build / tests (Markdown only). Format check clean.

https://claude.ai/code/session_01AmQ9QPAF1yskmQs7wSnZYc

---

Generated by Claude Code