KAYER CHECKOUT & ORDER OPERATIONS
One KAYER UA order flow from the Shopify cart through local payment and delivery to Dilovod accounting.
6+
core systems in one flow
2
payment paths: acquiring and invoice
4
Shopify order webhook topics
retry-safe
idempotency, retries and reconciliation
checkout.kayer.ua
loading…
KAYER Checkout handles the customer journey: it validates the cart, collects contact and delivery details, accepts online payment or a B2B invoice order and creates a consistent Shopify order. The order-operations backend maps SKUs to Dilovod product numbers, synchronises orders, prices and inventory, and gives operators a queue, error log and controlled recovery actions.
● If your situation is similar
CH-01C / ACTIONSend the current site, idea or blocker. I will tell you whether DIY, audit/rescue or a full build makes more sense.
Before / After
Was
- Payment, Shopify order, shipping and Dilovod document required separate checkpoints
- Repeated callbacks and webhooks created a risk of duplicate processing
- Diagnosis depended on logs across several services
Became
- One traceable route from cart to accounting document
- Signed callbacks, idempotency, queues, retries and reconciliation
- Checkout Admin and Order Operations /ops for daily operations and recovery
● REC — SYSTEM FLOW
Order-to-accounting flow
- 01Shopify cart
- 02KAYER Checkout
- 03LiqPay or B2B invoice
- 04Shopify order
- 05Order operations queue and worker
- 06Dilovod documents and inventory
automation-first · PL / UA / EN
● Project screenshots
Order-to-accounting architecture: Shopify cart, checkout, payment, order, operations queue and Dilovod.
What was wrong
The Ukrainian order process required LiqPay and Monobank, Nova Poshta delivery, a separate B2B invoice path, accounting documents and Dilovod synchronisation. Each stage needed a shared order identity, verified events and a visible recovery path.
What I built
I designed and built the external Next.js checkout, server-side cart validation, payment callbacks, Nova Poshta flow, B2B invoices and documents, Shopify order mapping and a protected operations panel. I also built the order-operations backend on FastAPI with HMAC-verified webhooks, a queue and worker, SKU mapping, order/price/stock synchronisation, retries, dead-letter states and an /ops dashboard.
Outcome
KAYER UA now has one operational route from cart to accounting. The normal flow carries order data into Dilovod, repeated events resolve through idempotency, and exceptions receive a status, audit record and controlled recovery action.
Next improvements
- Measure checkout conversion after a representative observation period
- Track the share of orders completed in Dilovod automatically
- Report median time from payment confirmation to accounting document
- Extend regression coverage for discounts, returns and complex B2B cases
● Metrics
Source tags show where each number comes from. Yellow badge = estimated, not verified.
core systems in one flow
operationspayment paths: acquiring and invoice
operationsShopify order webhook topics
operationsidempotency, retries and reconciliation
operationsMy role / internals
- Business-flow analysis and solution architecture
- Checkout UX for desktop and mobile
- Shopify API, App Proxy, webhooks and order mapping
- Payments, delivery, documents and bank reconciliation
- Dilovod integration, SKU mapping, queues and workers
- Security, observability, automated tests and deployment runbooks
Operations loop
- Customer flow: contact details, delivery and payment
- B2C acquiring and B2B invoice workflow
- Shopify order to operations queue to Dilovod sale order
- Price and inventory synchronisation with productNum to SKU mapping
- Admin and /ops tools for diagnosis and controlled recovery
Stack
Integrations
- Shopify Admin GraphQL API and four order webhook topics
- LiqPay, Monobank and bank reconciliation
- Nova Poshta and Checkbox
- Dilovod API and order-operations workers
- Supabase Storage, Resend and Telegram operations
Search visibility
- Public proof uses a synthetic checkout and protected login screens
- Search focus: Shopify Ukraine checkout, LiqPay, Nova Poshta and Dilovod integration
- Production operations panels remain private while the implementation is documented in the case
AI visibility
- Architecture and tests provide a structured context for safe agent-assisted maintenance
- Runbooks separate diagnosis from production actions
- Structured logs and explicit states give agents verifiable system context
Structured data
Schema.org types implemented with rich-result eligibility.
