The great abstraction: Why every leap in human progress looks the same

When we replace raw, messy, low-level substrate with a layer that represents context, productivity doesn't just improve incrementally. It explodes. Over the past few decades, modern software engineering has delivered staggering gains that would have been unimaginable to the engineers who came before. But why?

The reason isn't just better hardware or smarter people. It's that we built better abstractions. We stopped telling computers how to manage memory and started describing what we wanted to compute. The machine handled the rest.

Here's what's interesting: data engineering never got its equivalent moment.

The tools for data analytics have improved. Storage has become cheaper. Warehouses are becoming faster. But the fundamental workflow — the human effort required to transform raw data into a trusted business insight — has barely changed in decades.

Every new report, dashboard, or metric still demands heavy manual effort. Pipelines must be built. Logic must be reconciled across workflows. Definitions must remain consistent across a dozen different tools. 

All while teams run their own interpretation of what "revenue" or "active user" or "conversion rate" actually mean.

Meanwhile, the gap between what software teams and data teams can deliver has increased exponentially, slowing the speed at which organizations can turn data into trusted decisions.

Not because data engineers are missing skills.

It's because software engineering already has its abstraction layer, while data engineers are still waiting for one.

In other words, data engineering never received the abstraction breakthrough that transformed software engineering.

The key to unlocking that breakthrough is a universal semantic layer.

This is the core of Strategy Mosaic: a single, governed description of how the business works. It’s encoded once, and made available consistently to any tool, cloud, AI agent, or downstream system.

The same way high-level programming freed software engineers to focus on what they were building rather than how the machine executed it, a universal semantic layer frees data teams to focus on insight rather than reconciliation.

As a result, you define the business logic once, and run it everywhere.

That's what abstraction has always done. It removes complexity so people can focus on outcomes instead of mechanics.

And the productivity unlock that comes with it is just beginning.

Watch the full Freedom to Innovate keynote from Strategy World 2026.

If you'd like to find out more about the semantic layer, read our in-depth whitepaper about Strategy Mosaic.