Big Ideas in 2025: Enterprise + Fintech

When Regulation Becomes Code

✅ When Regulation Becomes Code

Companies in the banking, insurance, and healthcare industries spend countless hours and millions of dollars staying in compliance. Today, banking and insurance regulations span tens of thousands of pages; SBA lending documentation alone exceeds 1,000 pages. For businesses, keeping on top of these codes requires byzantine workflows and many hours spent hiring and training staff. Imagine, instead, that those lengthy documents — including text, images, and case precedents — could be used to train regulation-specific LLMs. Suddenly, compliance would become as simple as a Google query: “Is [X] compliant? What modifications need to be made?”

The onerous process of staying up to speed on regulation also poses a less obvious cost to consumers. To give just one example, an estimated 1.5 million consumers fall behind on their mortgages every year. What if those people could talk to someone steeped in Fannie Mae’s 1,000+ page servicing guide to get quick, accurate answers on how to modify their loan and get some relief? AI agents can be quickly trained and are infinitely patient. LLMs can streamline this traditionally fraught process.

The labor-intensive business of compliance is ripe for new software. AI can make our systems safer, more straightforward, and more efficient for consumers and companies.

✅ Ripping Out Legacy Systems of Record

AI is driving enterprise buyers to reconsider their entire tech stack. Klarna’s move earlier this year to replace Salesforce and Workday with homegrown, custom-built AI solutions is just the tip of the iceberg. I predict this sort of cord-cutting — ripping out legacy systems of record in favor of more dynamic upgrades — will be replicated many times over. For the first time in over a decade, systems of record are vulnerable. Whereas the prototypical software company of the 2010s plugged into existing systems of record to power downstream workflows, today the most ambitious founders are reimagining that core system entirely.

Relational databases will become multimodal: now that AI has advanced to actually performing work (instead of merely facilitating it), I expect to see customers seek out “systems of engagement” — dynamic, AI-powered tools that turn human “doers” into primarily reviewers. Systems of engagement will both store a core data set (customer details, order information, etc.) and serve as the primary application from which users complete their work.

This will not be fast or easy. The incumbents in these categories have deep data moats and vast amounts of resources.

Author

AiUTOMATING PEOPLE, ABN ASIA was founded by people with deep roots in academia, with work experience in the US, Holland, Hungary, Japan, South Korea, Singapore, and Vietnam. ABN Asia is where academia and technology meet opportunity. With our cutting-edge solutions and competent software development services, we're helping businesses level up and take on the global scene. Our commitment: Faster. Better. More reliable. In most cases: Cheaper as well.

Feel free to reach out to us whenever you require IT services, digital consulting, off-the-shelf software solutions, or if you'd like to send us requests for proposals (RFPs). You can contact us at [email protected]. We're ready to assist you with all your technology needs.

© ABN ASIA