Who decided I can delete a song from my Spotify playlist, but not from the artist's catalog?

The answer is business logic, i.e. the invisible rules that govern what we can and cannot do with data. But here's the twist: AI is about to flip this entire paradigm on its head.

The Anatomy of Modern Software: A Data Manipulation Story

Let's peek under the hood of any modern business application. Whether it's Spotify, Salesforce, or your company's inventory system, they all follow the same fundamental pattern.

Anatomy of modern software - Database (i.e. information) & Database Manipulators (i.e. applications)
A typical business software today consists two main conceptual sides.
1. A database to store the data.
2. An application to manipulate that data, such as creating, reading, updating, or deleting something (CRUD). The manipulation essentially forms the data into information.
We may see the applications in form of **backend**, containing the bulk of business logic. A **front-end** (with bit more business logic) for humans to interact with the backend, and an **API** for other systems to interface with the system.
But what's fascinating, is all these components exist for one primary purpose: To manipulate data according to predefined business rules.

The 'rule book'

While the intelligence lies with the users of the system, they are subjected to a set of restrictions, say 'rules' on how much they can have a free-hand and the capabilities using the system. These set of rules, i.e. business logic as governs your information management.

In a music streaming app's case, the rules might look like this:

Users can create playlists, Users can add songs to their playlists, Users can remove songs from their playlists
Users cannot delete songs from artist catalogs, Users cannot download songs without subscription

Where Intelligence Used to Live

Decision-making for data manipulation needs intelligence. The intelligence is served by the users while using the User Interface, or developers while writing the API client.

Traditionally, the intelligence in any software system resided with two entities: the developer who crafted the business logic, and the user who learned to navigate within those constraints.

The developer would spend months understanding customer needs, then encode those insights into rigid business rules. Users would then invest time learning how to work within these boundaries—clicking through interfaces, writing API calls, and adapting their workflows to fit the software's predetermined logic.

The AI Awakening: When Machines Understand Context

Then something remarkable happened. AI developed the ability to understand semantics—the meaning behind data, not just its structure.

Unlike traditional software that sees "Artist" as just a text field, AI understands that:

Artists create albums, Albums contain songs, Songs have genres, Users have preferences, Preferences drive listening behavior

This semantic understanding changes everything.

The reincarnation - The migration of grunt-work intelligence

We're witnessing a fundamental shift in where intelligence lives within software systems. The carefully crafted business logic that developers spent years perfecting? AI can now understand and apply it contextually, dynamically, and often more effectively than rigid code ever could.

Instead of users learning to navigate predetermined interfaces, AI agents can now:

Understand user intent from natural language
Navigate complex business logic automatically
Perform actions across multiple systems
Adapt to context without explicit programming

The Agentic Revolution: From User-Driven to AI-Driven

We are entering the age of agentic AI—where artificial intelligence doesn't just process requests, but actively performs tasks on behalf of users.

Consider this transformation:

Yesterday: User opens CRM → navigates to contacts → filters by criteria → exports data → opens email → composes message → sends to filtered contacts
Today: User says "Send our new product update to all enterprise clients who haven't purchased in 6 months"

The AI agent understands the intent, navigates the business logic, and executes the entire workflow.

P.S. Machine-to-machine protocols for 'semantic' communication

But it doesn't stop at human-to-AI interaction. We're seeing the emergence of new protocols like Model Context Protocol (MCP) that enable direct machine-to-machine communication based on intent rather than rigid API calls.

Instead of developers writing complex API integrations, AI systems can now communicate contextually:

"Find all customers similar to this one who might be interested in Product X"
"Optimize our inventory based on seasonal trends and current supply chain constraints"
"Reconcile financial data across all our systems and flag any discrepancies"

The New Data Manipulation Paradigm

AI has become the new manipulator of data. It understands context, applies business logic dynamically, and performs actions that previously required explicit programming for every scenario.

AI business-framework: The new data transformers

This doesn't mean traditional SaaS is dead—it means it's evolving into something more powerful. The rigid business logic we've carefully constructed over decades becomes the foundation that AI builds upon, not the ceiling that limits what's possible.

What This Means for SaaS

The software-as-a-service model is evolving rather than disappearing. Future SaaS applications won't just provide users with interfaces to work with data—they'll also provide AI systems with the tools, permissions, and business context to work with data on behalf of users.

The database, backend, and business logic remain important—they're still the foundation. But the interface layer is expanding to include both human-usable interfaces and AI-understandable protocols.

Looking Forward

SaaS isn't disappearing—it's adapting. The core value proposition stays the same: help businesses work with their data effectively. The mechanism is expanding from human-operated interfaces to include AI-powered agents.

Companies that succeed in this transition will be those that design systems for both human users and AI agents that may soon become common ways to interact with business software.

In the end, whether it's clicking "delete" on your playlist or asking an AI agent to optimize your business processes, we're all working with data in ways that create value. The tools are just getting more capable.

This transformation is already happening in many areas. The question for businesses is how to invest in the inevitable "AI Business-frameworks: The new data transformers (pun intended)" (Article coming soon).

Is SaaS Dead? No. Dying? No. Reincarnating? You bet.