Transforming Merchant Payment Processing

The first phase was mapping the full workflow for all three user types - merchants, processors, and administrators - and identifying where the current industry process breaks down. Merchants needed a guided onboarding flow that captured enough business context to generate meaningful matches. Processors needed lead management tools that matched how their sales teams actually work. Administrators needed visibility into the full platform without being bottlenecked in every transaction.

Design was done in Figma with a focus on role-specific clarity. Each user type sees a different interface on login - merchants see their application status and incoming offers, processors see their lead queue and offer templates, administrators see platform-wide activity. The design used a clean, professional visual language appropriate for a B2B financial product: high information density without visual noise, clear status indicators, and consistent action patterns across all three roles.

The frontend was built with React and Next.js. State management handled the complexity of a multi-role application where the same underlying data - a merchant application, a processor offer - renders differently depending on who is viewing it.

The backend was built on NestJS with Node.js, with PostgreSQL as the primary data store. AWS infrastructure and AWS Lambdas handled compute scaling. The integration layer connected three third-party services: OpenAI for AI-driven merchant categorization and offer analysis, Zoho and Zoho Sign for document management and e-signature workflows, and Retool for the internal admin dashboard.

Three distinct interfaces (merchants, processors, administrators); role assigned at authentication.

Structured application (business type, volume, industry) surfaces relevant processors and comparable offers.

Manage leads, create offers against applications, track deals to signed agreement; offer templates reduce repetitive quoting.

OpenAI classifies applications and provides contextual lead analysis.

Zoho Sign brings execution inside the platform; signed docs stored with the deal record automatically.

Instant updates on offer submissions, status changes, and document events.

A structured template system with defined fields (rates, fees, reserve %, contract length) plus free-text for processor-specific terms - comparable for merchants without over-constraining processors.

Keeping Payblox and Zoho Sign in sync required careful webhook handling and idempotent state transitions to avoid duplicate/inconsistent records.

Merchants can move from application submission to signed processor agreement entirely within Payblox - without managing separate conversations, email threads, or document tools. This was the core product promise, and the delivered build fulfills it.

The offer template system and lead management workflow replaced the spreadsheet-and-email process that processor sales teams were using. The Zoho Sign integration removed the document-chasing step entirely.

The OpenAI integration reduced the manual review time required to assess and categorize incoming merchant applications, allowing the matching process to move faster on the processor side.

The AWS infrastructure and TypeScript codebase give Payblox a platform that can handle growing volumes and add new features - additional processor types, new merchant categories, expanded AI capabilities - without rebuilding the core application.

Payblox plans to extend the platform with additional merchant categories, enhanced AI matching capabilities, and expanded processor tools as the marketplace grows.

The AWS Lambda architecture supports horizontal scaling as transaction volumes increase. The service adapter pattern in the integration layer makes adding new third-party tools straightforward without touching core application logic.