A jewelry-native storefront, a live inventory and marketplace for stones and SKUs, a built-in AI team that handles memos, repairs, restocks and reprices — and the connectors to plug into whatever ERP you already run. One platform that closes the gap between your management system and how the jewelry trade actually works.
Two ways in: spin up a brand-new jewelry website below, or .
Want to see houses before you build one? Browse live site examples →
From a digital studio launching its first collection to a manufacturer running a full bench — GemHub OS adapts to how the operation actually works. Same platform, different ways in.
Eight category surfaces, dozens of sub-modules behind each one — all sharing one database, one design system, and one team of AI agents.
Sales, stock, repairs, memos, agent decisions — live, drillable, and exportable to your board deck.
Multi-hub, multi-currency, multi-channel. 120,000+ stones, real-time movement, typed reservations.
Diamonds, jewelry, watches, melee — each track its own UX, all sharing your live catalogue.
Shopify, marketplace, wholesale, in-store, RFQ — one queue, one source of truth, one audit trail.
From drop-off to delivery, with customer comms, technician notes, and a clean P&L per ticket.
Track consignments at line-item level, with provenance, recall workflows, and aging alerts.
AI discovers the schema; a typed engine writes every row. Sandbox-first, audit-trailed, rollback-ready.
Per-role permissions, per-tenant data isolation, SAML/SSO, and a clean self-serve admin surface.
GemHub OS sits beside your ERP, your accounting system, and your POS — closing the jewelry-specific gap none of them were built to model.
Not a generic ERP with a jewelry skin. Four real workflows that don’t fit a standard OMS or accounting line-item — each one a native primitive in GemHub OS.
Every parcel out, every stone, with GIA certificate hashing — aged, recall-aware, reconciled against your GL.
The jewelry intelligence layer that knows the difference between a memo and a sale. Every capability below is LIVE in GemHub OS — backed by file evidence in 01b-ai-capability-inventory.md.
Drafts brand name, palette, voice, and a 12-SKU launch plan from a single prompt.
ERP gap resolved: Replaces the 6-week branding engagement small teams can’t afford.
Reads your legacy ERP / spreadsheet stack and proposes a typed mapping, asking the right business questions.
ERP gap resolved: ERP migrations stall on schema interrogation — GemHub OS does the asking.
Continuously monitors movement and flags stale memos, mispriced lots, and lost stones.
ERP gap resolved: The memo gap your ERP alone leaves open.
Checks GIA / EGL hash chain against registry and flags duplicate or modified certificates.
ERP gap resolved: Certificate fraud detection at parcel intake.
Watches gold / platinum spot and surfaces SKUs whose margin has crossed a threshold.
ERP gap resolved: Manual reprice cycles that lag spot moves by days.
Tracks sell-through against supplier lead time; recommends PO drafts with Nivoda / RapNet.
ERP gap resolved: Stockouts on bestsellers in fast-moving cuts and karats.
Flags memos > 60d / 90d with no movement and proposes recall or write-off batches.
ERP gap resolved: Stale consignments your ERP alone cannot age.
Computes labor + materials + bench-hours against quoted price; flags loss tickets in real time.
ERP gap resolved: Repair P&L invisible in standard accounting.
Expands a collection brief into a typed variant matrix with metal × karat × setting math.
ERP gap resolved: Combinatorial SKU explosion that overwhelms standard ERP.
Translates a row-level audit trail into a one-paragraph narrative for the auditor or board.
ERP gap resolved: Audit prep that consumed CFO days, not minutes.
Every ERP — whatever brand you run — was built to model dollars and SKUs. The jewelry trade lives one level deeper: parcels, certificates, repair tickets, consignments. GemHub OS closes the three places where that gap shows up.
Your ERP holds the GL but cannot track parcel-level memo-out, consignment-in, or per-stone certificate chain. Those live in spreadsheets and institutional memory — until they don’t.
ICP-T1A seasonal collection produces a combinatorial matrix of metal × karat × setting × stone grade × size. Standard ERP variant models were built for apparel, not stone-grade math.
ICP-T4A repair ticket carries a customer-owned item, a setting change, a quoted labor fee, and a photo trail. None of that fits a standard OMS line-item model.
ICP-T3| Capability | Your ERP alone | Generic SaaS point tools | GemHub OS |
|---|---|---|---|
| Parcel-level memo tracking | Manual / spreadsheet | Partial, no GL post | Per-stone, GL-posted |
| GIA certificate chain of custody | PDF attachment | Rare | Hashed, registry-aware |
| Repair ticket P&L | Workaround | External tool | Native, photo-trailed |
| Lot / stone-grade margin | Off-system | Apparel model | Native, per-stone |
Free during private beta. Production pricing scales with active SKUs and active modules — never with how many people log in.
* AI-11 and AI-12 are roadmap capabilities — not LIVE at private beta. Final pricing published before GA.
Join the US jewelry businesses building on the jewelry intelligence layer. Early access opens to a limited cohort each quarter.
A separate path for founders launching a new US jewelry brand — brand kit, storefront, and a curated collection structure, generated from a single prompt.
