Alekh Jindal is CEO and co-founder of Tursio, linking enterprise data to AI agents.
Not long ago, looking for public information meant finding the person who could get it for you. The concept goes back to the newspaper clipping services of the 1800s and the information brokers of the mid-20th century. It carried through to telephone reference services such as the New York Public Library’s Ask NYPL, where you can still call a librarian and ask them to look something up. There used to be an entire world of people whose job was to search for information on behalf of others.
Self-Serve Web Search
That changed at the turn of the 21st century with the rise of search engines. Vannevar Bush described the idea of information retrieval, the concept behind search engines, in his 1945 essay on the “Memex,” but it didn’t reach the mainstream until the 1990s, when crawlers and search engines like Lycos and Yahoo! became popular. Better ranking algorithms followed, like PageRank, and then Google came to dominate the field so completely that web searching became synonymous with “Googling.”
More recently, Perplexity challenged traditional link-based search with a natural language, LLM-powered Q&A interface, and Google even built its own “AI Overview” into search results. Web search made people self-serve, equipping them with the tools to find public information instead of relying on someone else to find it for them.
Self-Serve Data Search
Organizations today run on data, but access to it is often limited to a handful of experts. Everyone else relies on those people to get the data they need to do their work. Consider member outreach in a credit union: Someone needs to know who’s new and needs a welcome, or who needs attention because something’s going wrong.
That information exists in the core, in the CRM and in a handful of other systems. But the person who could act on it is usually waiting on someone else to find it, interpret it, reconcile it and hand it over, much like the information broker process of the mid-20th century. The question is whether this process can become self-serve, too, just like web search did, rather than people having to ask other people.
To be fair, business intelligence (BI) tools have been trying to make data self-serve for a couple of decades. The goal was to make enterprise data accessible to business users. Unfortunately, much of BI has turned into a dashboard-building exercise that still requires experts to build the models and dashboards, repeating the process over days or weeks with every new request. Ironically, BI tools have often made business users more dependent on experts, not less.
Build Vs. Buy
That raises the real question: How does an institution get there? Over the past couple of years, several organizations convinced themselves every business needs its own AI team and moved to build in-house. Some of those teams are now finding AI moves too quickly, leaving them slower and more expensive than expected. By the time an in-house build is operational, the outside world has usually moved on, and the team ends up reimplementing something that already exists, at a higher cost.
Credit unions and community banks are, in a way, better off here. Without the instinct to build everything in-house, they’re free to choose whatever is best out there instead of getting stuck in a loop of building their own. That’s a real advantage: staying nimble and switching to the best, most cost-effective option as the market moves, instead of being tied to whatever was built in-house years ago.
Closing The Gap
Self-serve data is likely to transition from an old myth to a new reality with AI. But the goal isn’t to build a search team from scratch. It’s to give the people who already know the members and the accounts the ability to find what they need without asking someone else to find it first. That capability is developing quickly, with a new wave of natural language interfaces designed to understand business users directly, without requiring a bank or credit union to own or maintain the underlying technology.
Web search used to be something you asked someone else to do. Now it’s something you do yourself without thinking twice. Data inside a financial institution hasn’t made that shift yet, but it will. The banks and credit unions that get there first will spend their time serving members instead of waiting in line for answers, and their members will feel the difference.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?