According to research by hybrid data platform provider Cloudera, 88% of IT leaders have added artificial intelligence to their workflows, with 60% using AI to improve the customer experience.
Specifically, they reported that AI helps them:
- Enhance security and fraud detection (59%)
- Automate aspects of customer support (58%)
- Leverage predictive customer service (57%)
- Power chatbots (55%)
Yet Cloudera chief strategy officer Abhas Ricky says financial services executives he speaks with often find it difficult to use AI to deliver a consistent customer experience across channels for a variety of reasons, including scattered data among different clouds and on-premises environments. Meanwhile, regulatory and security requirements make it difficult—if not impossible—for AI models to access this data at scale and on demand.
The best solution, he says, is a hybrid architecture that unifies all data for advanced analytics and data science while maintaining high levels of security and governance across environments. This approach enables financial institutions to leverage the power of AI without compromising trust.
“With hybrid cloud you can have an extremely flexible environment that ranges all the way from traditional [business intelligence] reporting to new, cutting-edge AI and machine learning applications,” he says.
Ahead, Ricky shares what a true hybrid model can enable for the financial sector, from better customer experience to flexible and scalable operations.
1. Flexibility And Scalability
Companies leverage the cloud for its flexibility and scalability. For most financial institutions, though, instant change isn’t always a selling point, as sensitive customer data cannot move between servers and networks without careful and methodical oversight.
Cloudera’s hybrid platform, with a consistent deployment model featuring robust security and governance, protects sensitive data while making it accessible and discoverable, reducing cloud risks and complexities. At its core is the open data lakehouse, blending the scalability and flexibility of data lakes with the governance, management and performance of data warehouses.
“The open data lakehouse enables large financial institutions to manage and analyze data at scale,” says Ricky. “You can do real-time analytics on both structured and unstructured data across any environment because there’s no data silos.”
All too often, collaboration requires data analysts and engineers to “meet in the middle,” so to speak—forcing them to work with tools designed to serve the most users at the expense of the highly specialized functions preferred by each party. Cloudera’s hybrid model promises flexibility at scale. “Data teams can collaborate or even work on their specific, preferred tools … across public and private clouds,” Ricky says.
The interoperability of hybrid architectures is more than just a convenience for the financial sector. It also reduces engineering costs, Ricky says. He recalls one executive at a large private bank telling him that refactoring a single application cost them $3 million. That’s not to mention the risks to business continuity during this process, he adds.
“Using a hybrid lakehouse, you can quickly scale to accommodate any growing data volumes without significant capital investments and … without application refactoring costs,” Ricky says. “[You can] build applications once, and deploy them anywhere.”
2. Unified, Secure And Compliant Data
A number of legislative frameworks already govern how organizations collect, store and analyze customer data, including the European Union’s General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA), to name just two.
With the widespread adoption of AI, regulators have additional concerns. In some jurisdictions, new legislation governs how AI accesses sensitive data, and how companies disclose the use of AI. The amended CCPA, for example, gives consumers more control over how organizations use their data to train AI models. There’s more AI legislation in front of Congress too, including the CREATE AI Act and the AI Research, Innovation and Accountability Act, which both have implications on AI development and deployment. Furthermore, lawmakers across the U.S. introduced almost 700 AI-related bills in 2024 alone.
Rather than waiting for legislative mandates, financial institutions must be more proactive in their data handling. Especially for banks that routinely work with an array of vendors, migrating to a hybrid cloud and open data lakehouse like Cloudera’s is the first step, says Ricky.
“You get a unified layer of security governance and compliance not only across multiple cloud and on-premises environments but also across multiple vendors,” he says.
Staying ahead of regulations can also drive better business outcomes. Ricky recalls Cloudera working with one large Indian bank where digital transformation paved the way for new consumer products. By consolidating their data in a secure, compliant hybrid cloud, he says the bank could finally analyze large volumes of data from multiple sources to yield better customer insights.
The result? Approximately two-thirds of the bank’s personal loans are now processed via database programs, and its machine learning-based credit models have achieved over 40% improvement in custom credit score models compared to standard models. Additionally, the bank is accelerating its value to end users too, adding nearly 250 features through a modern, user-friendly interface that has a ‘One View’ feature on the bank’s mobile app, allowing customers to manage their finances seamlessly.
“Using the Cloudera solutions, they were able to get not only multi-cloud flexibility and get the best levels of compliance for their regulators, but also significantly leapfrog their analytical capabilities and compete with the larger banks in the Indian financial institution landscape,” Ricky says.
3. Driving Innovation And Efficiency With AI And Gen AI
Just as large language models (LLMs) must be trained on enormous pools of text to appear as natural—and as human—as possible, banking AI models require troves of historical customer data to detect fraud, assess creditworthiness and perform other automated tasks. This data must be accurate and unbiased.
“We provide trusted data that is the core source for training and retraining models,” Ricky says. “You need to have high-quality, high-fidelity data sets.”
However, few financial decision-makers are eager to expose private customer data to a generative AI (gen AI) model, which stifles innovation. To assuage their fears, Ricky advises financial leaders to take greater ownership over their AI by placing its governance directly within R&D. This gives executives more visibility into—and control over—their gen AI’s training data access, lineage and integrity.
Migrating to a hybrid open data lakehouse makes this possible, Ricky says. “Unified data management … improves efficiency and enhances collaboration across teams and … gives everyone access to the same data sets, governed by the same rules.”
When banks take charge of their models and training data, they can build gen AI agents that strategically automate analyses and decision-making. At one bank, Ricky saw gen AI agents dramatically reduce the time needed for traders to start their day.
“Prior to the gen AI, you were spending a few hours reading documentation and synthesizing information,” Ricky says. “Now it takes two minutes, and you get 99% productivity improvements … That is where the world is headed in financial services.”
With secure hybrid cloud data platforms, financial sector leaders can usher their organizations into the next era, leveraging the latest technologies to achieve flexibility and scalability, optimize their data and ultimately better serve their customers.
