Search smarter, not harder with Intelligent Search

Firstly, tell us about the hypothesis behind Intelligent Search - why did you build it?

Bud has always been a leader in the financial data enrichment space thanks to our knowledge of machine learning and artificial intelligence. This means that we're able to take unstructured transaction data from our clients and enrich it, identifying things such as categories, merchants, processors, regularity and the location of the transactions. By doing this, we're adding a wealth of metadata to transactions that subsequently makes it easier for our customers to search and explore them. 

When you combine this with the knowledge that the “most common use of mobile banking is to check account balances or recent transactions” (95% percent of users) (Federal Reserve), it quickly became a no-brainer that we could use our experience in transaction enrichment to improve the customer experience of viewing and searching their transactions. 

By combining both Bud’s enrichment and generative AI capabilities, we’ve been able to build a best-in-class solution that allows customers to view and search through their transactions on their terms, whilst at the same time organically educating the customer about their finances.

What sort of ROI can a financial institution expect to get from Intelligent Search?

At every step of the way, Intelligent Search improves the user experience and, consequently, customer engagement. By getting these fundamentals right not only can financial institutions reassure customers that they know them, they can also delight them. According to Accenture’s 2025 report on consumer banking and advocacy, reassuring and delighting customers are two of the four drivers in creating customer advocates. And when a financial institution has advocates they can expect to see between a 5-30% boost in those customers’ share of wallet and see their overall revenue grow 1.7x faster than banks with the lowest customer advocacy scores. (Accenture)

We're also currently working on custom insights within Intelligent Search. This will allow clients to customise insights which are shown to customers within the context of what they're searching. Take for example a current account customer who is currently saving with a competitor. Clients will be able to configure Intelligent Search such that when customers are searching for savings transactions with competitors, the institution can include an insight about their saving account rates. This means that the expected ROI from Intelligent Search isn't just increased customer engagement and satisfaction, but it can also be used as an incredibly powerful tool to up-sell products to existing customers. 

‍

Despite only launching Intelligent Search earlier this year, you’ve already added new features to it. Can you tell us a bit about that?

One of the recent features that we’ve just added to Intelligent Search is our Smart Filters. These are shown after a customer has started searching and they allow customers to quickly identify all of the different reasons why transactions are being returned, whether that’s a match to the transaction description, the name of a merchant or even a similar category. 

These filters not only help customers to understand why we’re returning a set of transactions, but it also helps them to either filter down or refine their search at just the click of a button.

Take for example a customer searching for ‘Starbucks’. Smart Filters will recognise that the category ‘coffee’ is a common category amongst Starbucks transactions and, by clicking on the smart filter, customers can easily start viewing and understanding their entire coffee spend, not just that with Starbucks. This is an example of how Intelligent Search goes beyond just showing transactions to customers, but also helps them to organically learn about their spending habits.

Smart filters aren’t the only thing that we’ve been working on. By continuously iterating the capabilities of Intelligent Search, we aim to provide the best possible outcomes for our clients and their customers.

‍

Intelligent Search obviously makes use of generative AI. Can you talk about where this may fit into a financial institution’s wider AI strategy?

Intelligent Search is a great way for financial institutions to explore use of generative AI with their customers in a safe and secure way. Within Intelligent Search we use generative AI in two places. Firstly, in the wording of insights, we provide the LLM with the parameters and context of the insight from the customer's set of transactions and then allow the LLM to refine the wording. This ensures that insights continue to appear fresh and customers don’t get fatigued by repetitive wording.

Secondly, when a customer asks a direct question, such as: “How much did I spend on Starbucks this month?”, we allow the LLM to define what filters it should be using to search for the relevant transactions. In this example, that means recognising that ‘Starbucks’ is a merchant and the need to use a date filter. 

By constraining the LLM to specific tasks as well as separating it from the retrieval of transactions or conducting calculations, we can make use of the benefits of LLMs (such as their creativity in writing and dealing with ambiguity) while also reducing the risk of hallucinations. This combination makes it perfect for financial institutions who are eager to get started with generative AI in customer-facing applications.

Get in touch to find out how Intelligent Search can help you level up your digital banking experience today.