What if that future already happened and nobody noticed?
In 2013, seven PhD students at Berkeley built a tool to process huge amounts of data faster, focusing not on AI models, but on pure data engineering.
Thirteen years later, that company is worth 188 billion dollars [1]. It pulls in 6.9 billion dollars a year, growing 80% year over year [2]. More than 10,000 companies run their data on it [2].
This company is Databricks, and most people still think it's "just a data tool."
A Decade of Data Engineering Created an AI Advantage
Databricks spent its first decade doing something unglamorous: helping companies organize massive amounts of data in one place instead of scattered across ten different systems, but that work wasn't exciting, instead it was foundational groundwork.
Then generative AI showed up, and every company suddenly needed the same thing: a way to connect their own data to an AI model, safely, without leaking customer information or losing track of who accessed what.
Most companies didn't have that but Databricks already did. They had spent ten years building the governance, security, and data pipelines that AI now desperately needed. So instead of starting from zero, Databricks moved fast and in 2023 bought MosaicML, a company focused on training and fine tuning AI models, for 1.3 billion dollars [3]. That single move let Databricks go from "we store your data" to "we help you build AI on your data," almost overnight.
That's the real story. Databricks didn't get lucky. They spent a decade building the boring part, and when AI made the boring part valuable, they were the only ones already standing there.
Databricks Expertise Is Becoming a Critical Engineering Skill
If you work in data or AI, Databricks is quickly becoming what SQL or Git already are: a skill you're expected to have, not a nice extra on your resume.
More companies are standardizing their entire data and AI stack on one platform instead of stitching together five different tools. That means the engineers who understand that platform deeply become the ones trusted to build the systems everyone else depends on.
Enterprise AI Needs a Governed Data Foundation
Every company building AI faces the same risk: moving sensitive data to some new AI tool, and losing control over who sees it or how it's used.
Databricks solves that by keeping AI inside the same secure environment where the data already lives. No copying data to a separate system. No new security gaps to manage. Just one governed platform doing both jobs.
Certified Engineers Reduce Production Risk
Anyone can say "I know Databricks" on a resume, but passing a Databricks certification is a different story.
There are 7 official exams. Each one costs 200 dollars and you need 70% to pass. Even experienced engineers only pass on their first try about half the time [4]. That's not an easy badge to earn. It means the person actually knows how to build and fix real pipelines, not just talk about them in an interview.
For candidates, earning this credential proves mastery over real-world scenarios. From tuning complex Spark jobs to configuring multi-tenant governance in Unity Catalog under timed, proctored conditions. It signals to the market that they possess verified technical execution capability rather than superficial platform familiarity.
For enterprise customers, working with certified engineers eliminates trial-and-error overhead in production. It provides concrete assurance that the team handling their data infrastructure can optimize serverless compute costs, prevent governance vulnerabilities, and build high-throughput pipelines that scale reliably.
Factored's Databricks Expertise Is Independently Validated
We are Databricks partners, and we sponsored the Databricks Data + AI Summit this year.
And we've now earned proof of that expertise directly from Databricks. Factored just received the Databricks Brickbuilder Security & Governance Specialization, our first specialization [5]. Databricks doesn't hand this out for signing a partnership agreement. It requires certified engineers, real customer implementations, and for this specific track, a live technical assessment conducted in front of Databricks' own specialist teams.
So when a client needs to build or fix a data and AI platform, they're not starting with someone learning on the job. They're starting with a team Databricks itself has tested and validated.
Data really did become the new oil. AI really is the future. The only surprise is how long it took most of us to notice a company that bet on both, over a decade ago.
