The commerce operating system for modern retail
Power enterprise retail without limits.
Connect stores, inventory, orders, customers, AI and operations inside one enterprise-ready commerce platform.
Explore platformSOC 2 ready · 99.99% uptime · Four North American regions
0+
Storefronts live
$0B
GMV processed annually
0ms
Median API response
0
Largest chain, one tenant
Retail overviewLIVELast 30 daysAll channels
Net revenue
$18.42M
+12.4%
Orders
241,908
+8.1%
Sell-through
76.3%
+1.9%
Locations
312
4 opening
Revenue by channelRetail · Web · Marketplace · Wholesale
W18W20W22W24W26
Inventory signals
Denim · SoHoReorder 2,400
Outerwear · TorontoTransfer 860
Footwear · DC-WestOverstock 14%
Accessories · WebOn pace
AI forecast
Q3 demand up 9% in the Northeast. Three transfers drafted.
A platform enterprise retail can stand on.
Nine guarantees your architecture review will ask about, answered in the core product.
01
Enterprise security
SOC 2 controls, encryption everywhere, audited access.
02
Cloud native
Multi-region, containerised, zero-downtime deploys.
03
API first
Every capability in the console exists as an endpoint.
04
Low code
Ship workflows without waiting on a release train.
05
Scalable infrastructure
Peak traffic absorbed without a capacity conversation.
06
Role-based access
Granular permissions down to a single location.
07
Unlimited stores
Three locations or three thousand, one hierarchy.
08
High availability
99.99% contractual uptime with regional failover.
09
Global ready
Multi-currency, multi-tax, multi-language from day one.
01 — OUTCOMES
What changes in the first two quarters.
Aggregate movement across enterprise deployments in the United States and Canada, measured against the customer's own pre-migration baseline.
−34%
Order-to-fulfilment time
Routing resolves against one ledger instead of a nightly stock sync.
+18%
Sell-through on promoted lines
Transfers drafted before cover falls under the promotion window.
−61%
Integration maintenance hours
Point-to-point middleware retired in favour of the event bus.
11 wks
Median time to first live region
Pilot region carrying real traffic, running beside the legacy stack.
02 — COMMERCE PLATFORM
Nine modules. One data model.
Store operations
Staffing, tasks, cash and compliance for every location in one hierarchy.
Operations
Inventory intelligence
A single ledger across stores, warehouses and channels, reserved in real time.
Supply
Commerce automation
Routing, reorder, markdown and return rules that run without a nightly batch.
Automation
Customer management
One profile across POS, web and wholesale, with consent and segments attached.
CRM
Omnichannel
One cart, one customer, one order state — in store, online, marketplace, B2B.
Channels
Retail analytics
Sub-second dashboards on the transactional store — no warehouse round trip.
Analytics
Workflow builder
Compose triggers, conditions and actions visually. Export any workflow as code.
Low-code
AI assistant
Ask about stock, margin or a location in plain language; act on the answer.
Intelligence
Developer platform
REST and GraphQL parity, typed SDKs, a CLI and signed webhooks as standard.
Developers
03 — PRODUCT
Production screens, not marketing renders.
Five surfaces from the console. Every number, state and control is what an operator actually sees.
Retail analyticsCohort: all channels · 90 daysStreaming
Conversion
3.84%
AOV
$164.20
Return rate
6.1%
Repeat rate
41.7%
Revenue trend
This yearLast year
Top locations
New York · SoHo$2.14M
Toronto · Yorkdale$1.86M
Chicago · Michigan Ave$1.62M
Boston · Newbury$1.20M
Seattle · Pike$0.98M
04 — SOLUTIONS
Built for how your business actually sells.
Retail chains
Central control with regional autonomy across hundreds of locations.
312 stores, one tenant
Franchise groups
Shared catalog and brand rules, separate ledgers and permissions.
Per-franchisee scoping
Luxury retail
Clienteling, appointments and made-to-order lines held to inventory.
Appointment-aware stock
Wholesale
Price books, terms, credit limits and bulk order approval flows.
B2B terms engine
Department stores
Concessions, consignment and vendor settlement in one order model.
Vendor settlement
B2B commerce
Buyer portals, quotes and repeat ordering on the same engine as retail.
Quote to order
B2C commerce
Storefront APIs, subscriptions and returns tied to the store network.
Ship-from-store
05 — IMPLEMENTATION
A migration with no freeze window.
Potmain runs alongside your existing stack while traffic moves channel by channel. Nothing is switched off until the replacement has been carrying load for a full cycle.
WEEKS 1–2
Discovery
Channel inventory, data model mapping and a written migration order.
Solutions architect
WEEKS 3–5
Foundation
Tenant, regions, RBAC and the connectors your stack already depends on.
Platform team
WEEKS 6–10
Pilot region
One region live on real traffic beside the legacy stack, with rollback intact.
Joint
WEEKS 11–18
Rollout
Remaining regions and channels, sequenced around your trading calendar.
Your team
ONGOING
Optimise
Quarterly architecture reviews, workflow tuning and forecast retraining.
Named TAM
06 — ARCHITECTURE
One engine. Everything connected.
Composable services on a single data model. Adopt one module or run the whole stack — the contract never changes.
LAYER 01Retail channelsStores · Web · Mobile · Marketplace · Wholesale portal
LAYER 02Potmain Commerce EngineEvent busUnified data modelPolicy & RBAC
Inventory
Single ledger, reservations
Orders
State machine, splits, returns
POS
Offline-tolerant terminals
ERP
SAP, NetSuite, Dynamics
CRM
Unified customer profile
Warehouse
WMS, wave picking
Payments
Tokenised, multi-PSP
Marketplace
Amazon, Walmart, eBay
Shipping
Rates, labels, tracking
Analytics
Streaming, warehouse sync
AI
Forecasts, agents, tools
07 — INTELLIGENCE
AI across every retail workflow.
Not a copilot bolted on the side. Forecasting, pricing and replenishment run inside the engine, on your live operational data.
Inventory forecasting
Per-SKU, per-location demand curves that update as the day sells.
Demand prediction
Seasonality, promotions and weather resolved into one horizon.
Store intelligence
Where a location is under-indexed against comparable stores, and why.
Sales insights
Channel-level projections with confidence intervals, not point guesses.
Automation
Agents draft transfers, reorders and markdowns for human approval.
Recommendations
Assortment and pricing moves ranked by margin impact.
Natural-language analytics
Ask in plain language; get the query, the answer and the action.
Agent-ready architecture
Tool-callable APIs and scoped credentials for autonomous systems.
08 — DEVELOPERSOpen developer portal
An API your team will actually enjoy.
Typed SDKs, idempotent writes, signed webhooks, a CLI, and a GraphQL surface that mirrors REST one-to-one.
REST & GraphQL
Full parity, one schema
SDKs
TS · Python · Go · Java
Webhooks
Signed, replayable
Authentication
OAuth 2.1, scoped keys
CLI
Deploy workflows from CI
Sandboxes
Per-branch environments
api.potmain.com/v4
POST /v4/orders
{
"channel": "retail",
"location": "loc_soho_01",
"currency": "USD",
"lines": [
{ "sku": "DNM-501-32", "qty": 2 },
{ "sku": "ACC-BLT-M", "qty": 1 }
],
"fulfillment": { "strategy": "nearest_stock" }
}
→ 201 Created
{ "id": "ord_9f2b41", "state": "reserved" }200 OK38msidempotency-key: ord_9f2b
09 — SECURITY
Enterprise security, by default.
Controls your security team can audit on day one — not a roadmap item.
SOC 2
Type II controls, evidence on request
SSO & SCIM
Okta, Entra, Google
RBAC
Role inheritance per location
Encryption
AES-256 at rest, TLS 1.3
Audit logs
Every write, exportable
Compliance
PCI DSS, CCPA, PIPEDA
Monitoring
Live SLOs and status feed
Cloud infrastructure
Isolated tenants, four regions
10 — CUSTOMER STORIES
−34%
We replaced four systems with one engine and cut order-to-fulfilment time by a third.
VP, Retail Technology
National specialty chain · 480 stores
11 wks
The first platform our engineers asked to keep after the pilot.
Chief Information Officer
Franchise group · US & Canada
3×
Peak weekend traffic tripled. We did not open a capacity ticket.
Director of Digital Commerce
Omnichannel wholesaler
Illustrative examples. Named references available under NDA.
Where Potmain runs.
Four North American regions with tenant isolation, and enterprise delivery teams in the metros where our customers operate.
us-east-1N. VirginiaPrimary · US East
us-central-1DallasUS Central
us-west-2OregonUS West
ca-central-1MontréalCanadian residency
Enterprise retail teams served
New YorkTorontoChicagoLos AngelesDallasSeattleAtlantaBostonVancouverMontréalMiamiDenverMinneapolisPhiladelphiaCalgarySan FranciscoPhoenixOttawa
11 — RESOURCES
Field notes for retail technology teams.
EVALUATION GUIDEComposable commerce for retail chains: a CIO evaluation checklistThe twenty-two questions worth asking a commerce vendor before an architecture review.14 min readARCHITECTUREWhy a single inventory ledger beats nightly synchronisationWhat breaks when three systems each believe they own stock, and how one ledger fixes it.9 min readOPERATIONSShip-from-store economics in North AmericaLabour cost per pick, split-shipment penalties and where the margin actually lands.11 min readMIGRATIONReplacing legacy commerce without a freeze windowSequencing channels so the old stack is retired only after the new one carries load.12 min readSECURITYRBAC patterns for multi-banner retail groupsModelling franchisees, concessions and regional autonomy without cloning tenants.8 min readAIForecasting demand with promotion and weather signalsHow per-location demand curves are trained, evaluated and safely acted on.10 min read
12 — DEFINITIONS
What is a commerce operating system?
A commerce operating system is a single platform that owns the shared data model for retail — inventory, orders, customers, locations and pricing — and exposes every operation as an API and an event. Channels and back-office systems read one source of truth instead of synchronising copies of it.
Unified data model
One definition of a product, a location, an order and a customer, shared by every channel and every integration.
Event bus
Every state change is published as a signed, replayable event, so downstream systems react instead of polling.
Composable adoption
Modules can be adopted one at a time. The API contract does not change as you take on more of the platform.
Agent-ready architecture
Tool-callable endpoints and scoped credentials let autonomous systems act inside the same permission model as people.
Questions from enterprise buyers.
01How long does an enterprise migration take?+
Most retailers run Potmain alongside their existing stack for one quarter, moving channel by channel. A 300-location rollout typically completes in 14 to 20 weeks with a dedicated solutions architect.
02Can Potmain run alongside our ERP?+
Yes. Potmain is the commerce layer, not a replacement for finance. Certified connectors exist for SAP, NetSuite and Dynamics, and the event bus handles anything custom.
03What are the data residency options?+
Four North American regions with tenant isolation. Data stays in the region you select, including backups and analytics replicas.
04How is pricing structured?+
An annual platform fee by tier, plus usage on orders processed. No per-seat charges and no revenue share.
05Do you support B2B and B2C on one instance?+
One instance, one catalog, separate price books, terms and approval flows. Wholesale portals share the same order model as retail.
06How do you handle peak season?+
Capacity is elastic and load-tested against your own traffic shape before Black Friday. There is no capacity ticket and no surcharge.
07What does support look like?+
A named technical account manager, 24/7 severity-one response with a 15-minute target, and quarterly architecture reviews.
08Does Potmain replace our POS hardware?+
No. The terminal application runs on the hardware you already have — iOS, Android and Windows terminals, with certified payment devices from the major North American processors.
09What happens when a store loses connectivity?+
Terminals keep selling. Transactions queue locally with the last known price and stock position, then reconcile against the ledger when the connection returns. Conflicts surface as a review queue, not a silent overwrite.
10How do you handle customisation without upgrade debt?+
Extensions run outside the core in your own sandbox environments, and workflows are versioned artefacts deployed from CI. Nothing you build blocks a platform upgrade.
11Can we export our data if we leave?+
Yes. Full schema export in CSV or Parquet, plus a documented dump of the event log. There is no proprietary format and no exit fee.
Build the future of retail commerce.
A 30-minute technical walkthrough with a solutions architect. No slides.