|
|
|
|
August 21, 2025
|
Hackers Infiltrate Alleged North Korean Operative’s Computer, Leak Evidence of...
|
|
August 21, 2025
|
Ecosia Proposes Unusual Stewardship Model for Google Chrome
|
|
August 21, 2025
|
OpenAI Presses Meta for Evidence on Musk’s $97 Billion Takeover Bid
|
|
August 15, 2025
|
ChatGPT Mobile App Surpasses $2 Billion in Consumer Spending, Dominating Rivals
|
|
|
How Brex Rebuilt Its AI Procurement Process to Keep Up With the Pace of Innovation
July 6, 2025
As artificial intelligence tools flood the market at breakneck speed, companies are struggling to keep up — not just with adoption, but with identifying what’s worth their time in the first place. For fast-moving organizations trying to stay competitive, the traditional enterprise procurement cycle is becoming more of a liability than a safeguard.
Brex, the fintech unicorn known for its corporate credit card and expense management platform, ran into this very issue head-on — and instead of sticking with what wasn’t working, it tore up the rulebook.
Speaking at the HumanX AI Conference in March, Brex CTO James Reggio offered an inside look at how the company radically revamped the way it evaluates and adopts AI tools.
When the Process Breaks the Product
Like many companies, Brex initially tried to assess new AI tools using its existing procurement process: structured, thorough, and slow. That playbook had worked well in the past — but the rules changed in the wake of ChatGPT’s explosive debut.
“In the first year following ChatGPT, when all these new tools were coming on the scene, the procurement process itself would actually run so long that the teams that were asking to procure a tool lost interest in the tool by the time that we actually got through all of the necessary internal controls,” Reggio said.
By the time internal approvals cleared, teams had either found other solutions — or the AI tools themselves had evolved so much that they no longer met the need.
Rethinking Procurement for the AI Era
Realizing that long, cautious pilots were now a risk — not a safeguard — Brex opted for a new approach. It started with a foundational change: redesigning the legal and data processing framework that governs how third-party tools are brought into the company.
“We built a new framework for data processing agreements and legal validations,” Reggio explained. The idea was to speed up legal and compliance review without compromising on security or privacy — a delicate balance, but a necessary one.
This helped the company test tools more quickly and move from interest to evaluation while the internal demand and market momentum were still alive.
Letting Teams Lead the Way
Brex didn’t stop at legal streamlining. It flipped the traditional top-down software approval model by empowering engineers and end users to drive tool adoption. At the heart of this was a concept they call the “superhuman product-market-fit test.”
Rather than relying on leadership to decide which AI tools are worth scaling, Brex turned to the people actually using them. “We go deep with the folks who are getting the most value out of the tool to figure out whether it is actually unique enough to retain,” Reggio said.
If a tool is beloved by a team and meaningfully improves productivity, it passes the test. If not, it’s quietly phased out. “We’ve definitely canceled and not renewed maybe five to ten different larger deployments,” he added.
A $50 Budget — and a Lot of Autonomy
One of the most distinctive parts of Brex’s new approach: each engineer gets a $50/month budget to license approved tools of their choice.
“By delegating that spending authority to the individuals who are going to be leveraging this, they make the optimal decisions for optimizing their workflows,” Reggio said. This decentralized model allows Brex to experiment broadly, without waiting on slow top-down decisions or locking into expensive licensing deals too early.
And it’s working: “We haven’t seen a convergence,” Reggio noted. That is, engineers haven’t all rushed to the same tool (like Cursor, for example), which suggests true experimentation and fit-finding is happening at the individual level.
This micro-budgeting strategy also helps inform macro purchasing. Once usage hits a critical mass, Brex can negotiate larger enterprise licenses with a better understanding of real demand.
Embracing the Messiness of AI Adoption
If there’s one piece of advice Reggio had for other enterprises facing the same AI overwhelm, it’s this: don’t wait for perfect clarity.
“Knowing that you’re not going to always make the right decision out of the gate is just paramount to making sure that you don’t get left behind,” he said. “The one mistake that we could make is to overthink this and spend six to nine months evaluating everything very carefully before we deploy it. And you don’t know what the world is going to look like nine months from now.”
In a space where product roadmaps are rewritten monthly and competitors are experimenting in public, speed matters more than certainty. Brex is betting that the companies that win in AI aren’t the ones with the most rigorous procurement playbooks — but the ones who are willing to break them entirely.
|
|
|
Sign Up to
Our Newsletter!
Get the latest news in tech.
|
|
|
