NCR Voyix Pay — Worldpay Payment Processing Integration

Overview
Starting in August 2025 at NCR Voyix, I served as lead engineer for the webhook system and a key contributor to the overall architecture of Voyix Pay’s Worldpay payment processing integration. Voyix Pay uses Worldpay as the authorization gateway, then processes the approved transactions and funds the merchants — positioning NCR Voyix as the payment orchestrator between merchants and the global payment rails.
Voyix Pay integrates across all of NCR Voyix’s major POS platforms — Aloha for restaurants, Counterpoint for grocery, and others in their retail portfolio. The Worldpay integration I helped build handles millions of transactions per day, representing billions of dollars in annual payment volume. This is not a side project. It is the critical path for all incoming transaction processing.
The integration was designed as a distributed, container- and VM-backed Python system. I led the webhook team, participated in all technical decision-making meetings, and was a key voice in validating architectural and infrastructure choices across the full stack.
Architecture Overview
The payment processing pipeline has three major moving parts:
Webhook Service (FastAPI)
My primary ownership. This service receives real-time payment events from Worldpay as transactions are authorized, settled, or reversed. Built with FastAPI, it handles event ingestion, validation, and routing to downstream processing systems. The webhook layer is the first point of truth — if this fails, nothing downstream works.
Processor Service (REST API Client)
Worldpay’s APIs return paginated batch and transaction data. The processor pulls thousands of pages of API responses — batching, settlement data, transaction status updates — and feeds them into the pipeline. I contributed to the processor’s design, including the pagination strategy, error handling, and backoff logic needed to reliably consume massive datasets from Worldpay without timeouts or data loss.
Data Pipeline (Oracle + PL/SQL)
Incoming transaction data is staged, sanitized, cleansed, and enriched before it reaches downstream systems. This pipeline uses Oracle as the database backend with PL/SQL stored procedures for the heavy lifting: data validation, transformation, merchant enrichment, and readiness checks before settlement. The PL/SQL layer ensures that only valid, clean data reaches the merchant funding systems — because dirty data in payment processing means lost money.
Key Challenges
Scale and Data Volume
The processor service pulls thousands of paginated API responses from Worldpay across batch settlement windows. Building a retrieval system that can handle this volume without overwhelming either Worldpay’s rate limits or our internal systems required careful pacing, parallelization strategy, and robust retry logic.
Data Quality at Scale
With millions of daily transactions, data issues compound fast. The PL/SQL pipeline had to handle real-world edge cases: mismatched currency formats, invalid merchant identifiers, partial settlement records, timezone discrepancies, and duplicate events. The sanitization and enrichment layer is what separates a payment system that works in testing from one that works in production.
Infrastructure Flexibility
The system uses a mix of containers and VMs depending on workload characteristics. Some components benefit from container orchestration and rapid scaling, while others (like certain database or integration services) needed VM-level stability. Designing for a hybrid deployment model added complexity but gave us the flexibility to meet performance requirements.
Real-Time vs Batch Processing
The webhook handles real-time events, while the processor handles batch settlement data. Reconciling these two pipelines — ensuring real-time authorizations match batch settlement records — is critical for accurate merchant funding. Any discrepancy means money problems.
Technical Approach
- Python/FastAPI: Webhook service for real-time Worldpay event ingestion
- Python REST clients: Processor service for paginated API data retrieval from Worldpay
- Oracle Database + PL/SQL: Data staging, sanitization, cleansing, and enrichment pipeline
- Container + VM hybrid: Flexible deployment model matching infrastructure to workload needs
- Idempotent processing: Ensures no duplicate transactions even under retry scenarios
- Comprehensive error handling: Worldpay response codes mapped to internal taxonomies for meaningful downstream routing
Business Impact
- Critical-path integration: All incoming Voyix Pay transactions flow through this system — millions per day, billions in annual volume
- Merchant funding accuracy: Clean, validated data ensures merchants are funded correctly, building trust with NCR’s retail and hospitality clients
- POS platform coverage: Integration spans Aloha (restaurants), Counterpoint (grocery), and NCR’s broader retail POS portfolio
- Operational reliability: Robust error handling and data validation reduce payment processing failures and reconciliation issues
- Worldpay as authorizer: Positions Worldpay’s globally-established payment rails as the authorization layer, with NCR Voyix controlling the post-authorization processing and merchant funding flow
Technologies
- Python (FastAPI, REST API clients)
- Worldpay Card Payments API and Payments API
- Oracle Database with PL/SQL
- Container orchestration and VM deployment
- Webhook architecture and event-driven design
- Paginated API consumption at scale
- PCI DSS compliance frameworks
- Enterprise payment processing architecture
Partnership Context
The Worldpay integration was part of a strategic partnership between NCR Voyix and Worldpay announced in early 2025. The collaboration integrated Worldpay’s acquiring services into NCR Voyix’s cloud-based commerce platform, delivering a unified point-of-sale and payment processing solution for retailers and restaurants. The partnership leveraged Worldpay’s presence in over 146 countries and support for 135 currencies, giving NCR Voyix’s merchants global payment capabilities from Day One.
This move also positioned NCR Voyix competitively against bundled rivals like Toast, Square, and Shift4 — companies that successfully combine payment processing with software. By integrating Worldpay, NCR Voyix transitioned from a pure POS software vendor to a full commerce+payments platform, creating higher customer retention through increased switching costs and stable, recurring payment revenue.
Lessons Learned
At the scale Voyix Pay operates, there is no such thing as an “edge case” — everything is a normal case when you’re processing millions of transactions a day. The data pipeline isn’t a luxury, it’s a necessity. And the engineers who build payment systems need to understand that every bug has a dollar sign attached to it.
This project also reinforced the value of hybrid infrastructure thinking. Containers are great, VMs are great, but the best engineers know when to use each. The system works because we matched the right infrastructure to the right workload, not because we dogmatically picked one model over the other.
Resources
- NCR Voyix & Worldpay Partnership Analysis — Third-party coverage of the partnership and its strategic impact