AI Will Reshape SaaS. Systems of Record Aren’t Going Away.

Lately, the conversation has been about how AI is killing SaaS. I have no doubt that AI will reshape how software is built, priced and delivered. 

But, I don’t expect us to be replacing our finance systems of records with AI-generated code anytime soon. 

Boards and investors are asking about AI, teams are experimenting with it and the pressure to move fast is real. 

AI can accelerate workflows and improve decision support. Replacing a system of record like an ERP or quote-to-cash system is a different question. These systems embed decades of domain expertise, regulatory nuance, compliance requirements, and edge cases that rarely show up in demos.

That depth exists for a reason. Getting it wrong carries real domain and compliance risk. It also raises a practical question: do you really want your engineers rebuilding infrastructure that already reflects years of accumulated expertise?

At the OTTO-MATES AI Strategy Summit, I shared a simple point that continues to resonate: in finance, 75% right is not good enough.

The reason is simple. Finance operates in a dynamic, highly regulated environment where accuracy is non-negotiable.

Complexity Is Compounding

Revenue is not getting any simpler. From usage-based and hybrid pricing models to global billing across dozens of jurisdictions and evolving accounting standards, something changes every month. 

A true system of record absorbs that complexity and connects billing, revenue, tax, and reporting into a single source of truth.

Replacing that with a general AI model assumes the model fully understands every workflow, exception, and every change in real time.

That assumption deserves scrutiny.

Why Finance Is Different

More people are getting comfortable with self-driving cars. There’s still something comforting about knowing that if something goes wrong, the car can pull over. 

But how many would board a commercial airplane with no pilot?

The difference is recoverability.

In some environments, mistakes are manageable. In others, the consequences compound immediately.

Revenue systems determine how you report earnings and comply with regulation. There is no safe place to pull over if they are wrong. 

Finance systems of record look a lot more like the airplane.

When you sign the 10-Q, there is no AI disclaimer. 

Where AI Creates Leverage

The right question isn’t whether AI replaces your system of record but where AI creates leverage without increasing risk.

Here are three practical guidelines:

• Protect the Core. Don’t experiment with the integrity of your general ledger, revenue engine, or compliance framework.
• Target the Friction. Apply AI to repetitive, data-heavy workflows that sit on top of the core, such as collections prioritization, invoice routing, or exception flagging, or operational analytics. These are areas where speed matters, but outcomes remain anchored to the system of record.
• Demand Explainability. If an AI-driven workflow impacts financial outcomes, you should be able to trace how a decision was made. Involve your auditors early so there are no surprises later. If you can’t explain it to an auditor or your board, you shouldn’t rely on it. 
•  

Don’t rip out the foundation. Instead layer intelligence on top of trusted infrastructure and drive measurable efficiency. Grow revenue without growing headcount at the same rate.

The leaders who win here won’t chase every demo. They will protect the core and apply AI where it creates measurable leverage.

The more interesting debate ahead isn’t whether AI replaces enterprise software, but instead if finance leaders will rethink the model from build versus buy to buy first, then build intelligently on top.