AI Will Revolutionize Traditional Enterprise Software
In discussions about the coming disruption caused by AI, the spotlight often falls on consumer tech. Let’s discuss how AI is coming for one of Silicon Valley’s biggest cash cows: Enterprise software.
Enterprise Software Breakdown
Like the name, enterprise software isn’t really fancy. If you break it down, mostly it comes down to four things:
A data model.
A user experience.
Business rules.
Analytics.
Let's dissect each of these components:
Data Model
An enterprise data model is largely a commodity at this point. Business users can often sketch it out on a whiteboard without relying on expensive enterprise software packages. Some organizations may already operate in this manner, perhaps with costly vendor consultants handling customizations. When you cut that hefty check to Oracle or SAP, you're not primarily paying for the data model.
User Experience
Speaking of things that aren’t a value-add in the enterprise software world, have you seen the user experience of most of these packages? Even with so-called “modern” enterprise SaaS, the UI and UX are typically a train wreck. So much so that a lot of organizations pay their own developers a lot of money to build 100%-custom interfaces on top of the vendor’s APIs (if they exist) or database (if they don’t).
User experience is near and dear to my heart, and I’ll do another post at some point about what I think AI means for its future. But suffice it to say, it’s quite likely that AI-driven interface generators will vastly outperform the kinds of UI builders that enterprise vendors have been shipping since the Windows 3.1 days in terms of ease of use and enjoyment.
Business Rules
This is where your substantial investment in enterprise software begins to pay off. There’s a lot of complexity in the modern enterprise, and a lot of value in not having IT code and debug it all themselves. Even something we all feel like we implicitly understand — say, HR onboarding and offboarding — is full of exceptions and conditions. And watch out when you get into areas like contracting and compliance.
The thing is, business rules align perfectly with what modern AI can do. If your enterprise’s business rules can be well-documented in consistent and plain language, it seems like a short leap to feed them to an LLM with a prompt to automate them with a database backend, APIs, and a UI. Yes, it’s not going to be 100% fidelity out of the box (yet!) but 80-90% is within reach. And from there, your IT and business analysts can fine-tune and validate the solution, providing a more tailored and cost-effective alternative to off-the-shelf or customized vendor solutions.
Analytics
The analytics world was undergoing dramatic shifts even before LLMs burst onto the scene. But one of the biggest trends is toward true enterprise-wide analytics that span all of the systems and datasets in the organization. In this world, analytics provided by any one software package become a lot less important and, indeed, a lot less powerful.
As others have observed, AI is ideally suited to the business analytics space, and all of the modern analytics tools that were already encroaching on traditional enterprise analytics are rapidly incorporating cool new AI features.
Conclusion
What does this mean if you’re a product or technology leader in the enterprise software space? As always, the right solution is to embrace these trends and look for opportunities to AI-enable your product offerings. Provide AI-chat-based UIs. Use AI for no-code data model and business rule customizations. Integrate your system with modern AI analytics. But don’t wait: Unlike a lot of the AI hype, in the enterprise software space there’s little holding AI back and the change is going to happen fast!
