Wtf is Neomir Cortex?

Let's go back to 2019.

At the time, I was working as a master data manager. My job was to ensure that SAP data was consistent with data from another system. Sounds simple - until you actually try to do it. Neomir didn’t exist back then. And neither did a data quality tool that could handle cross-system checks the way I needed.

So I did what everyone would have done:

I opened Microsoft Excel :-)

What followed were VLOOKUP formulas so complex they probably violated several laws of nature. The solution was clunky, but it did the trick and kiiindaaa worked. My past self was happy because I could do cross-system data checks.

Untiiiill... The datasets grew beyond a couple hundred rows. That was when Excel crashed.

The Challenge

Cross-system data quality is fundamentally hard.

You need to:

• Pull data from multiple systems

• Join it efficiently

• Iterate quickly while refining rules

Most tools fail at one of these steps. Excel collapses under volume. Databases are powerful, but your data isn't centralized there. Data warehouses are hard to set up.

This is the gap Neomir Cortex was built to close.

What Neomir Cortex is

Neomir Cortex is:

A high-performance analytics engine.

You can think of it as a tiny, short-lived data warehouse - designed not for storage, but for speed. Cortex loads data from different systems, executes joins, filters, and aggregations at high performance, and then discards everything once the rule has been validated. No persistence. No unnecessary overhead.

That design choice is intentional.

Data quality work is inherently exploratory:

• You define a rule
• Run it
• Inspect the results
• Adjust and rerun

Cortex is optimized for exactly that and thereby keeps the feedback loop tight.

What it evolved into

As we built more connectors, another problem became obvious: Many cloud applications don’t support full SQL (e.g. Salesforce that can't perform joins). That limitation makes meaningful cross-table validations nearly impossible.

Cortex sidesteps these constraints. By extracting and processing data locally in a purpose-built engine, Neomir enables complex validations even when source systems themselves are limited.

Final thoughts

Neomir Cortex is not a feature.

It’s a foundational technology - the engine that already powers cross-system validations today and will enable many more capabilities in the future.

Most data quality problems don’t fail because teams don’t care. They fail because the tooling was never built for the real complexity of modern data landscapes.

Cortex exists because I ran into that wall myself - and decided it shouldn’t exist anymore.