Enriched Data at the Heart of the Digital Banking Revolution | BNP Paribas, Kroo, Snowdrop | The Fintech Show #147

The next revolution in digital banking will come from enriched data. 

In the latest episode of The Fintech Show we’re joined by Ken Hart from Snowdrop Solutions, Waleran Guinard from BNP Paribas and Alexey Gabsatarov from Kroo Bank to look at how AI-driven transaction enrichment is transforming banking.

The cleaning and enriching of transaction data has the potential to enhance customer trust and bring about more personalised banking services. Tune in for the latest on how banks can utilise this, avoid errors and much more.

The Problem: Confusion in Banking Transactions

Many are discussing how AI, as big of a topic as it is, can actually be brought in to enhance banking. Some estimates of the value it could bring to the sector reach the heights of $1 Trillion.

Solving the problem of unclear transaction data is an area that could immediately benefit. Customers encountering cryptic merchant names or confusing details on their statements can cause unnecessary anxiety, lead to disputes, and overwhelm customer support systems. For example, a transaction might display the legal name of a franchise instead of the familiar brand name, making it difficult for users to recall the purchase.

This lack of clarity results in mistrust and frustration. The problem for banks, Guinard, who is Group Head of Customer Cards notes, is that disputes over such transactions waste both customer and bank resources. So resolving this problem is massive for banks.

The Solution: Advanced Transaction Enrichment

In the episode, CEO Ken Hart describes how Snowdrop Solutions address this challenge through a sophisticated API that leverages AI to clean and enrich transaction data. By processing over 1.4 billion messy transactions monthly with 98-99% accuracy, the solution transforms ambiguous data into clear, user-friendly formats. This includes accurate merchant names, precise locations, and additional attributes like the type of establishment (e.g., “romantic Italian restaurant”).

Guinard also elaborates on how BNP Paribas have incorporated these solutions, going beyond basic name corrections. They integrate enriched data, such as logos, Google Maps locations, customer reviews, contact details, and merchant websites, directly into their digital platforms. This approach enhances usability and fosters a deeper connection with customers.

Building Trust Through Quality Data

According to Alexey Gabsatarov who is CTO at Kroo, the foundation of effective AI models lies in the quality of data. So by partnering with Snowdrop, Kroo Bank has seen measurable improvements: a 15% increase in customer engagement and a noticeable reduction in complaints. Clean, enriched data not only resolves ambiguities but also instils confidence in users, encouraging them to trust and engage more with their bank.

Hart further explained the emotional benefits of clarity, noting that removing uncertainty reduces customer anxiety. When users feel in control of their finances, they’re more likely to explore additional banking products, deepening their relationship with the institution.

Customization and Accuracy With AI

The speakers also delved into the technical aspects of how AI is powering these solutions. Of course, with all the hype, we all want to know how AI will actually change the customer experience. 

Gabsatarov outlined the importance of training models on focused, high-quality datasets to avoid irrelevant or erroneous outputs. Unlike generic large language models, specialised AI systems are tailored to deliver precise results, such as ensuring correct business names or categorising merchant types.

Hart addressed the problem of AI “hallucinations,” where models generate plausible but incorrect information. Snowdrop counters this issue by grounding their tools in real-world data, such as Google Places’ database of over 200 million verified locations. This ensures that enriched transaction data remains accurate and trustworthy.

Unlocking New Possibilities

Beyond clarity, enriched transaction data opens up opportunities for personalization and convenience. Hart and Guinard shared examples of how this technology can enhance banking services: 

• Travel Features which automatically suggest multi-currency accounts or loyalty programs for users travelling abroad.  
• Expense Insights providing detailed spending breakdowns (e.g., “How much did I spend at Amazon in February?”) through AI-powered queries.  
• There is also the possibility for memorable experiences which help users recall past visits to specific merchants, complete with contact details and reviews, to make repeat interactions seamless.

There are further insights in this great episode, so be sure to watch the full thing above. And catch more fintech content just like this, on our website.