Banks across the world have spent years testing artificial intelligence, yet many still struggle to move beyond pilot projects. Plumery AI believes it has found a way to close that gap. The digital banking platform has launched a new product, called AI Fabric, designed to help financial institutions embed AI into everyday operations without weakening governance, security, or regulatory controls.
The company describes AI Fabric as a standardised integration layer that connects generative AI tools directly to core banking data and services. Rather than relying on one-off, custom-built integrations, the framework uses an event-driven, API-first architecture intended to scale as banks expand their digital capabilities.
Turning AI experiments into production systems
The challenge Plumery is addressing is well known in the financial sector. While banks have invested heavily in AI research and proofs of concept, relatively few initiatives make it into full production. Consulting firms such as McKinsey have repeatedly highlighted the same obstacles: fragmented data environments, legacy systems, and operating models that were not designed for continuous AI deployment.
Team DigitalDefynd
According to Plumery, AI Fabric aims to solve this by providing shared infrastructure and governed data streams that can be reused across multiple use cases. The idea is to give banks a consistent foundation for AI, rather than adding new tools on top of already complex systems.
Ben Goldin, Plumery’s founder and chief executive, said banks are clear about their priorities when it comes to AI adoption. They want solutions that deliver measurable improvements in customer experience and operational efficiency, but not at the expense of control or compliance. In his view, an event-driven data architecture changes how banking data is produced and shared, making AI easier to operationalise rather than harder to manage.
Fragmented data remains a core obstacle
Data fragmentation continues to be one of the biggest barriers to AI at scale. Many banks operate legacy core systems alongside newer digital channels, resulting in siloed customer data and disconnected product journeys. Each new AI initiative often requires fresh integration work, separate security assessments, and additional governance reviews, all of which slow progress and increase costs.
Research into explainable AI in financial services supports this concern. Fragmented data pipelines make it harder to trace decisions and raise regulatory risk, particularly in sensitive areas such as credit scoring and anti-money-laundering. Regulators have been clear that banks must be able to explain and audit AI-driven outcomes, regardless of how or where the models are built.
Plumery says its approach addresses these issues by presenting banking data as domain-oriented, governed streams that can be safely reused. By separating systems of record from systems of engagement and intelligence, the company argues that banks can innovate more confidently while maintaining oversight.
AI already plays a role in banking
Despite the structural challenges, AI is already in active use across the financial sector. Large banks routinely deploy machine learning and natural language processing in customer service, risk management, and compliance.
Citibank, for example, uses AI-powered chatbots to handle routine customer enquiries, easing pressure on call centres. Other institutions rely on predictive analytics to monitor loan portfolios and anticipate potential defaults. Santander has publicly outlined how machine learning supports its credit risk assessments and portfolio management.
Fraud detection is another area where AI is firmly established. By analysing transaction patterns in real time, AI systems can identify suspicious behaviour more effectively than traditional rule-based approaches. However, industry analysts note that these models depend on high-quality, well-integrated data flows, something smaller institutions often struggle to achieve.
More advanced applications, such as conversational AI for advisory services, are emerging more cautiously. Academic research into large language models suggests they could support certain banking functions under strict governance, but such use cases remain closely scrutinised due to regulatory and ethical considerations.
Xin Jie Chua
Platform-based approaches gain momentum
Plumery is entering a competitive market of digital banking platforms that position themselves as orchestration layers rather than replacements for core systems. Its partnerships, including an integration with open banking provider Ozone API, reflect a broader push toward ecosystem-based solutions that reduce custom development.
Dan Brightmore
This strategy aligns with a wider industry shift toward composable architectures. Vendors such as Backbase promote API-centric platforms that allow banks to plug in AI, analytics, and third-party services as needed. Analysts generally agree that this incremental approach is more practical than large-scale core system replacement.
Uneven readiness across the sector
Readiness for large-scale AI adoption remains uneven. Research from Boston Consulting Group suggests that fewer than one in four banks believe they are prepared for enterprise-wide AI deployment. Governance frameworks, data foundations, and operational discipline continue to lag behind ambition.
Regulators have responded by offering controlled environments for experimentation. In the UK and elsewhere, regulatory sandboxes allow banks to test AI-driven solutions under supervision, encouraging innovation while reinforcing accountability.
Kalyeena Makortoff
For vendors like Plumery, the opportunity lies in bridging the gap between technological ambition and regulatory reality. AI Fabric enters a market where demand for operational AI is clear, but where success depends on demonstrating transparency, safety, and control.
Whether Plumery’s framework becomes widely adopted remains to be seen. What is clear is that as banks shift their focus from experimentation to execution, the underlying architecture supporting AI will matter as much as the models themselves. Platforms that can balance flexibility with governance are likely to shape the next phase of digital banking.
