Financial Data Integration

Enterprise data architecture and automation for the Truist / SunTrust / BB&T merger

Truist Financial Corporation � SunTrust Banks � BB&T Corporation (via Infosys) � May 2019 � Jul 2023

Context

The BB&T / SunTrust merger � announced February 2019 and closed December 2019, forming Truist Financial � was one of the largest financial institution mergers in recent US history. The combined entity serves approximately 10 million households across roughly 2,000 branch locations. The data integration program required consolidating large-scale, heterogeneous core-banking systems under tight regulatory timelines, with no room for data quality failures.

This engagement ran across the SunTrust, BB&T, and Truist phases of the program from May 2019 to July 2023, encompassing pre-merger preparation, active integration, and post-merger stabilization.

Work

Data Architecture Analysis & Integration Design

Led data architecture analysis and integration design across legacy SunTrust and BB&T systems � source system profiling, data mapping, preprocessing specification, integration testing, and issue resolution. Produced architecture diagrams and mapping documents that served as the authoritative reference for integration teams.

End-to-End Pipeline Automation

Designed and implemented automated data pipeline architectures supporting consolidation across the merged data estate. Automation frameworks streamlined data reconciliation across heterogeneous core-banking source systems, significantly reducing manual intervention in data-movement and validation workflows. These pipelines formed a foundational layer of the integration infrastructure and were critical to meeting data-readiness milestones for regulatory and operational go-live.

Dependency-Aware Batch Validation

Engineered a dependency-aware batch validation framework (Python) for Mainframe/Unix pipelines. The framework enforced sequencing guarantees across complex pipeline graphs and provided structured, actionable error reporting � enabling teams to identify and resolve integration failures before they propagated downstream.

ML Risk Classification (NDMP)

Built a machine learning classification model within the Non-Default Management Platform (NDMP) workstream for loan default prediction. This represented an early, production deployment of predictive modeling in a live financial-services environment, demonstrating the integration of ML capabilities with enterprise data governance requirements.

Data Quality & Analytics

Developed analytics and data-quality monitoring workflows that enabled detection of integration anomalies and validation of completeness across the merged data estate � essential for both regulatory compliance and operational readiness. Automated batch log analysis and line-of-business insight visualization.

Recognition

Sindhu J. Manuvalil, Software Engineering Director at Truist Financial, on this engagement:

“He was not executing tasks � he was architecting solutions. When the program encountered technically complex integration scenarios, Mr. Pokhrel was consistently the resource we turned to for resolution. What distinguished him from other technical contributors was not only his depth of skill but the independence and judgment with which he operated.”

Prior: Fidelity Investments (May 2018 � Mar 2019)

As Principal Automation Architect at Fidelity Investments (via Infosys), led enhancement of the ETL automation framework � reviewed architectural limitations, drove improvements into team delivery, validated source systems and fact/dimension warehouse targets, and monitored Informatica PowerCenter workflows for data lineage and quality assurance.