Project Status
Last updated: 2026-05-18.
Read these first, in order
docs/brief/concept-narrative.md— current articulation of the product concept and the layered-model design thesis.docs/discovery/customer-zero-profile.md— GMI as the design partner / first test surface.docs/v0/scope.md— v0 build scope.- This document.
- The rest of
docs/as needed.
Who works on this product
- Tyler Georgiadis — builder, repo owner. Sole engineer at present.
- Greg Sher — product co-lead, sales / industry. Co-founder of GMI. Industry expertise and connections are a primary design input and the basis for the eventual sales motion.
This repo is the shared product design brain for both contributors.
Current stage
v0 build planning complete; ready for platform-repo handoff. v0 scope memo reviewed by Greg (no industry-side blind spots flagged) and locked. Three synthetic supplier profiles and two fictional retailer profiles are in place as v0 test surfaces, sitting under docs/v0/ alongside the scope memo. APIC competitive posture verified by attending the supplier-onboarding meeting. Implementation work moves to the platform repo from here; this repo continues to hold product strategy, discovery, and v0-scope artifacts.
What has been decided
- Layered canonical + transformation model is the load-bearing design thesis. Shared canonical layer (categories, facets, primary fields) + per-retailer transformation layer (configurable mappings producing retailer-shaped outputs). See
docs/decisions/0001-layered-canonical-model.md. - Sweetwater is the design reference — best-in-class US music retailer with ~300+ standardised facets per category. Their public taxonomy/facet structure anchors v0 design.
- Target market: APIC/MusiPOS users + smaller retailers without their own solution. Not the largest national chains with mature in-house data ops.
- Cross-retailer supplier overlap (H2) is confirmed by industry knowledge: AU brands have single exclusive distributors, so retailers stocking the same brand share the same supplier. Shared normalised layer is the structurally correct shape.
- Generalisability beyond instruments is confirmed. The design isn't instrument-specific.
- Greg Sher is product co-lead and the primary sales / industry driver. GMI is customer zero — co-designer, sanity-checker, and likely reference customer (not "the wedge target").
- APIC is not a meaningful competitor at present. See
docs/landscape/apic.md. - Design quality is a v0 priority, not a v1 polish concern. UX is enforced from the first commit — visual quality, interaction polish, and a shared design system across admin and retailer surfaces. See "Design quality bar" in
docs/v0/scope.md.
What's open
- H3 supplier participation in a centralised platform (individual-retailer behaviour is positive; platform-level still unproven). APIC supplier meeting attended week of 2026-05-12; it produced competitive verification but not H3-shaped supplier-behaviour intel — see
docs/landscape/apic.md. Next signal will come from direct supplier conversations rather than APIC observation. - H4 pricing / commercial model. Deferred until v0 demonstrates value.
- H5 transformation layer absorbs retailer divergence cleanly — this is the design hypothesis v0 proves; v0 build is the test.
- Multi-retailer interview sprint (
docs/discovery/customer-discovery-plan.md) — not yet scheduled; lower priority now given H1/H2 confirmed.
What's stale or partial in this repo
docs/discovery/customer-discovery-plan.md— was authored when H1/H2 were open; some assumption thresholds have been overtaken. Still useful as a framework for the eventual multi-retailer sprint.docs/discovery/open-questions.md— partially answered; some questions superseded.docs/discovery/assumptions-and-hypotheses.md— original assumption log, mostly superseded by the concept narrative and the risk-ranked version in the customer discovery plan.docs/brief/product-brief.md— original inception brief; the MVP it describes is overtaken by the v0 scope memo.docs/brief/concept-narrative.mdis the current articulation.
Recently completed (pre-handoff)
- v0 scope memo at
docs/v0/scope.md— reviewed by Greg, no industry-side blind spots raised, locked as the handoff brief. - Structured workflow walkthrough call with Greg — held. Output folded into
docs/discovery/gmi-walkthrough-notes.mdrather than a separate map doc. - APIC supplier-onboarding meeting (week of 2026-05-12) — attended. Findings folded into
docs/landscape/apic.md. Reinforced "not a meaningful competitor" verdict; did not produce H3-shaped supplier-behaviour intel. - Synthetic supplier profile set — three deliberately-varied synthetic suppliers (A/B/C) drafted by 2026-05-18; all live under
docs/v0/. The C profile forced two new v0 commitments captured indocs/v0/scope.mdunder Per-supplier mapping config: v0 capabilities: (a) a missing-column warning at ingest (not full schema-drift detection — v0 expects operator awareness when a supplier's file shape changes; the platform flags missing saved-mapping columns and otherwise leaves drift to operator re-mapping); (b) per-supplier (or per-file-variant) missing-product handling with two v0 modes — ignore-missing-rows, snapshot-diff. Operator-confirm-per-push and grace-period are both deferred post-v0; suppliers like C with mixed snapshot/delta pushes and no status column default to snapshot-diff with known misclassification — v0 resolution is to escalate to the supplier rather than build per-push operator gating. - Retailer profiles for the v0 transformation targets — JBM and CMS at
docs/v0/joe-bloggs-music-profile.mdanddocs/v0/capital-music-sound-profile.md; GMI partner profile atdocs/discovery/customer-zero-profile.md.
Active next steps
- Platform repo handoff — implementation conversation moves there with
docs/v0/scope.mdas the brief. This is the next concrete action; the rest of this list is parallel/optional. - Curation workstream — define the v0 facet set per category, draw canonical values from supplier data + Sweetwater reference (
research/inputs/sweetwater-category-reference-2026-05-13.xlsx). Content work, parallel to software. Not a handoff blocker; required for credible demo. - Direct supplier conversations — to produce H3 signal that the APIC meeting did not. Lower urgency than handoff but the open H3 path.
Repo map
CLAUDE.md— repo behaviour rules and stage awareness.STATUS.md— this file.docs/brief/product-brief.md— original inception brief (banner).concept-narrative.md— current concept articulation (load-bearing).docs/decisions/0001-layered-canonical-model.md— the load-bearing design thesis.docs/v0/scope.md— v0 build scope memo (load-bearing for platform-repo handoff).supplier-a-profile.md— synthetic rich-data supplier profile (v0 ingestion test surface).supplier-b-profile.md— synthetic medium supplier profile, print music distributor archetype (v0 ingestion test surface; covers Print Music v0-priority category).supplier-c-profile.md— synthetic degraded supplier profile, drum-and-percussion wholesaler archetype (v0 ingestion test surface; covers Drums v0-priority category; exercises cross-supplier identity resolution and forces schema-drift + deletion-detection v0 commitments).joe-bloggs-music-profile.md— fictional small-retailer profile (transformation-layer test surface).capital-music-sound-profile.md— Cin7-shape retailer profile (output-schema stress case).docs/discovery/customer-zero-profile.md— GMI as the design partner. Also referenced as one of the v0 retailer transformation targets; the GMI partner relationship lives here, the v0 fixture role is referenced fromdocs/v0/scope.md.gmi-walkthrough-notes.md— captured workflow knowledge from Greg's accounts.customer-discovery-plan.md— multi-retailer plan (lower priority now).assumptions-and-hypotheses.md— original assumption log (banner).open-questions.md— refreshed open questions.docs/landscape/apic.md— competitive landscape: APIC.apic-database-reference.md— technical reference for APIC's database (carried over).docs/templates/decision-memo-template.mdfictional-retailer-profile-template.mdsupplier-source-profile-template.mddocs/api/,docs/roadmap/,docs/taxonomy/— empty, intentionally.research/inputs/— partner-shared raw data, local-only (seeresearch/inputs/README.md)..claude/rules/— repo rule files referenced by CLAUDE.md.