Everyone is still talking about AI.

But the more conversations I have with engineering teams, the less they talk about AI… and the more they talk about something else entirely.

Quiet rebuilds.

Not new products

Not big launches

Not transformation programmes

Just teams going back and fixing what’s already there

Pipelines that have grown messy over time

Data models that no one fully trusts anymore

Dashboards that mean different things to different teams

Workflows held together by a mix of legacy logic and good intentions

It’s not exciting work on paper, but it’s becoming a priority

Because without it, everything else slows down

AI doesn’t land properly

Analytics becomes inconsistent

Decision making drifts

So instead of pushing forward, a lot of teams are stepping sideways

Refactoring

Simplifying

Standardising

Doing the kind of data engineering work that rarely gets talked about publicly, but makes everything else possible

The interesting bit is this

The companies doing this well aren’t necessarily the ones shouting the loudest about AI

They’re the ones quietly investing in solid data engineering capability and giving those teams the time to do it properly

If you’re in that phase right now, you’re definitely not alone

And if you’re trying to hire people who can actually stabilise and improve data platforms (not just build shiny new things), that’s where the market is getting really tight