APAC Banking
Most Banks Are Fixing The Wrong AI Decisions
The highest value AI opportunities are often hidden inside operational transactions with the greatest economic exposure, execution frequency, and downstream impact.
Transactional AI
The architecture problem is now the bottleneck
The industry is converging back toward core transaction platforms for real time AI execution.
AI Economics
The highest AI ROI happens inside the transaction
The highest value AI decisions occur at the point where transactions and economic impact intersect.
Industry Domains
When You Can Only Inspect 3 Percent
Border agencies cannot inspect every shipment. The real challenge is maximizing inspection yield through better targeting, risk intelligence, and transactional analysis.
Background
About James Roca
With over three decades of global IT experience, James advises enterprises on AI strategy, transactional systems, fraud intelligence, and operational transformation.
Insights
Ideas shaping Enterprise AI
Original analysis on enterprise AI strategy, economics, decisioning, architecture and adoption. Exploring where AI creates value, how it changes business decisions, what it takes to reach production, and the forces shaping how enterprises invest in and deploy AI.
insights
AI Does Not Transform Businesses. Better Decisions Do
Enterprise AI adoption is accelerating, yet the economic evidence is far less decisive. Most CEOs say their AI transformations are realizing only part of their intended results, and widespread adoption has produced little profit and loss impact. The usual explanations point to scale, data, talent or change management. The deeper error is mistaking the deployment of AI for the transformation of the business. AI does not transform a business by generating content or predicting an outcome. Better decisions do.
insights · 10 min read
AI Value Decays with Distance from the Decision
The same intelligence can have radically different economic value depending on when and where it arrives. A payer that finds an unsupported claim before payment can hold it, while the same finding after payment only starts a recovery. A bank that sees financial stress before a missed payment can help, while the same insight after default is a collections signal. What changes is the organization's remaining ability to influence the outcome. AI creates the most value when it arrives while there are still practical options, and that value decays with distance from the decision.
insights · 11 min read
Every Payment Provider Has Transaction Data. Few Have Merchant Intelligence.
Payment providers can describe a merchant's transactions with extraordinary precision and still misunderstand the business behind them. They know what was authorized, settled, refunded and disputed. They often cannot tell whether a merchant is genuinely growing, shifting volume to a competitor, approaching a liquidity constraint or likely to benefit from an intervention. AI is closing that gap, but only when transaction records are connected to merchant context, converted into decision-specific evidence and measured against outcomes.
insights · 11 min read
From Laboratory Automation to Intelligent Diagnostic Operations
For decades, diagnostic laboratories competed through scale, automation, network reach and payer access. AI does not make those strengths obsolete. It creates a second basis of competition through the ability to coordinate specimen, information, operational and financial decisions across the laboratory journey. The next advantage is intelligent diagnostic operations.
Strategic approach
AI Strategy
Most AI programmes underdeliver because they start with capability and work backward to value. These articles argue for a different starting point: identifying where AI investment produces the highest measurable return, building governance that functions at production speed, and sustaining model performance over time.
Framework
Scaling AI Adoption Starts With Understanding the Business
Enterprises have invested heavily in the capability to run AI. Far fewer have developed the capability to determine where AI should be applied.
Explore
Framework
What If Every Client Conversation Began With the Knowledge of Your Best Industry Expert?
Most enterprises have invested in making information available. Far fewer have made expert judgment reusable.
Explore
ai strategy
The Fraud Solution Opportunity Waiting on IBM Z
IBM has spent years building the ability to run AI inside one of the world's most important transaction-processing environments. Telum and Telum II put model inference close to the applications and data on which consequential financial decisions depend. That is a substantial engineering achievement, but it is not yet the full commercial opportunity. The runtime makes it possible to execute models in the transaction path. The greater prize is the fraud solution built around it, the models, behavioral features, decision logic and workflows that let institutions prevent more fraud rather than merely run more inference.
ExploreCore principle
Intelligence at the core of every decision
AI strategy isn't about technology adoption—it's about embedding intelligence into the decision-making fabric of your organization. Every transaction, every risk assessment, every customer interaction becomes an opportunity to learn, adapt, and improve. The question isn't whether to use AI, but how to architect your organization so intelligence flows to where decisions are made.
Learn about our approachIndustry focus
Industry Domains
AI strategy applied in specific operational contexts. These articles examine how Transactional AI, governance, and decision economics work in practice across payments networks, financial services, insurance, and the transaction systems at the core of global commerce.
Government Border Control
Operational Intelligence for Border Control: How AI on IBM Z Strengthens the Decisions That Protect a Nation
Border agencies face a structural and permanent challenge — more decisions, fewer people, higher stakes. The data to make better decisions already exists in the transaction systems agencies trust. The constraint has always been architectural: analytics ran somewhere else, on a different timeline, against a copy of the data. IBM Z's Telum processor changes that. For the first time, AI inference runs inside the transaction itself — against the authoritative record, at the moment the decision is made, before the window closes.
Payments Networks
Acceptance Gaps: The Network Volume That Left and Never Came Back
Acceptance gaps represent permanent lost volume. Spending that flows to cash, bank transfer, or a competing network because card acceptance is unavailable, unreliable, or uneconomic in that merchant segment does not automatically return when the acceptance gap is closed. Closing gaps earlier is more valuable than closing them later. Knowing which gaps to close first requires analysis that aggregate acceptance data cannot provide.
Payments Networks
Acquirer Risk: The Failure You Did Not See Coming Is the Most Expensive Kind
A single large acquirer failure can expose a payment network to hundreds of millions in unrecovered settlement losses. The network sees every transaction flowing through every acquirer's portfolio — far more granular signal than any credit agency can provide. Early detection while mitigation options are still available is the investment that makes the most consequential network risk manageable.