One Platform, One Profile, AI-Powered Execution

The P2P platform is the consumer side of the global payments company's two-platform fintech architecture, covering peer-to-peer payments, the social feed, merchant checkout through pay via the platform, cards and balance, and crypto. Taller's engagement at the P2P platform and the global payments company organization is the deepest fintech engineering footprint in the portfolio: multi-year, multi-pod, spanning the consumer side at the P2P platform, the platform side at this company, and the processor side at their payment processing platform (still active in 2026). The five cases below describe what Taller has delivered against the strategic surfaces the P2P platform's report proposed.

The global payments company's One Platform consolidation strategy (their in-person payments platform → the P2P platform → their payment processing platform → their unified payments platform common chassis) and the AI execution program (AI unit-test generation, Java 17 migration acceleration, compliance investigation collapse) require platform primitives delivered against the consolidation architecture and AI/ML engineering capacity supporting the broader program.

Taller delivered two platform-consolidation primitives. The Post-Purchase Notifications component integrates with the global payments company's AI-powered cashback and Smart Receipts products, with the legacy C++ migrated to Java on the latest their internal API framework. The Unified Native Onboarding platform consolidated thirty-plus disparate onboarding flows into one adaptable, localizable experience. Same-client AI engagement runs through contributor presence on the global payments company's Agent Builder — the self-service agent platform — plus Machine Learning Engineer roles supporting the global payments company's large-scale ML infrastructure (recommendation, ads ranking, personalization, search, and NLP on Spark, EMR, and Kafka).

The Post-Purchase Notifications delivery shipped AI-powered personalized notifications, modernized the legacy codebase to Java and enhanced system performance. The Unified Native Onboarding platform consolidated thirty-plus flows into one platform with context-driven UX and global localization.

The results from the global payments company's AI execution program are concrete: the checkout app moved from weekly to daily and multi-daily releases, AI-generated unit tests delivered in minutes instead of two to three sprints, Java 17 migrations from thirty hours to five, ten-to-thirty percent reductions in coding effort, and compliance investigations collapsing from days to minutes. These are the global payments company program outcomes spanning more workstreams than any one engineering function owns. The consolidation strategy and the AI execution program belong to the global payments company. The two consolidation primitives Taller delivered, plus contributor capacity into the Agent Builder team and the broader ML infrastructure, sit underneath that program.