Formula 1: Engineering When Every Millisecond Counts
ANAHEIM, CALIF. — You could argue collaboration tooling company Atlassian sponsored a Formula 1 team because it wanted broader brand recognition. Or because it’s a robust metaphor for modern software engineering.
While both of these are true, the sport has evolved into a feat of engineering itself, making Atlassian Williams Racing a fascinating collaboration, not just a brand deal.
Formula 1 is one of the most complex use cases of software engineering, data science, rapid feature development, agility, resiliency, incident response, testing and distributed team collaboration on Earth.
While most F1 fans focus on the team’s two drivers and maybe the team principal James Vowles, the Atlassian Williams Racing (AWR) team actually has about 1,100 members.
“If they’re all pointing in completely different directions, you’re done. There is zero chance that you’ll be fighting for wins or championships,” said Vowles on the keynote stage at Atlassian Team ’25.
“Gone are the days where one individual can make a difference by what they produce. The only way you bring together 1,000 people pointing all the right way is technology. The next element of things is everything is wrapped in data.”
It’s no surprise that the Atlassian Williams Racing partnership was a key part of this week’s Team ’25. What is surprising is what the pairing has learned in the three months since it was announced. And like all things tech, it comes down to the people.
Rapid Prototyping and Testing
In Formula 1, most sprints last less than five days, with the first beta release test drive on Friday, alpha qualifying rounds on Saturday and full general availability at the Grand Prix on Sunday. F1 leaves the pressure of most engineering sprints in the dust.
“You want to go with the development of the car long term, but also short term for that race weekend,” said Jenson Button, former Williams driver and current AWR ambassador, in the same keynote chat.
With every lap, terabytes of data is generated, which is then analyzed by a whole team of data scientists back in the Oxford headquarters.
“They need to be very precise in what they do,” Andrew Boyagi, head of DevOps evangelism at Atlassian, said in a press briefing. “To watch that collaboration unfold in real time, and how important it is that they’re very precise in the way that they work together, it’s pretty incredible.”
Each year, a new car is built essentially from scratch, which generates more data, including documentation around design, factory outputs and any race-day incidents that can happen anywhere from the garage to the practice laps to the event itself.
“You’ve got all this information that’s generated by very different crafts that all need to work together to get to one outcome,” Boyagi said. “Certainly something like Rovo with search and being able to find information easily is definitely going to be very helpful,” he said, referring to Atlassian’s AI-powered platform built on top of its Teamwork Graph, which had its general availability announced during Team ’25.
Not only is this data powerful, it’s challenging, as Formula 1 races occur across 21 countries on five continents, each with its own data residency rules.
All at 330 kilometers per hour. With design and build kicking off a year and a half a race.
“The cars are treated as prototypes. We are trying to adapt the cars every single race,” Vowles said. “We’re trying to bring air dynamic upgrades, or suspension upgrades, or any number of other systems upgrades to the car every single race, in order to maintain our competitiveness relative to the field.”
This year’s car could’ve easily won the end of last year’s tournament, he said. But so could the cars of any of the other nine racing teams, all evolving rapidly at the same time. The Atlassian Williams Racing team learned things in Suzuka, Japan’s event last weekend that it is actively applying to Bahrain’s race this weekend, implementing updates to the front wing and the floor that were conceived six to eight weeks ago.
“That’s what I love about our business. An idea that someone has a conversation is on the car a few weeks later,” Vowles said. “But again, everyone’s doing that. What you’re trying to do is make sure each idea you have has more lap time efficiency than someone else,” as each racing team aims to slightly edge their way up the field via continuous delivery of new features.
Each car has more than 20,000 components. This translates to 50,000 channels of data flowing across borders, with a combination of AI and data streams and very real human drivers in the loop.
One clarification that Vowles made is that Formula 1 is inherently not Big Data, because each track is so different.
“Big Data is when you have a billion transactions with some sort of boundaries to it,” he said. “We’re changing the cars every track, and our competitors are changing the cars every track. The tires change every track.
“You’ve really got to be very careful about getting your consistent baseline through that you can learn from.”
Human-in-the-loop drivers are what drives data consistency, the keynote speakers agreed, as the most expensive but also the best sensor any car has.
Working With Real Deadlines
Formula 1 remains a huge example for resiliency and backup plans.
Within the stringent budgets and regulations that govern F1 teams, the AWR Oxford headquarters is always manufacturing parts, both for next year’s car and this year’s iterations, plus spare parts.
The tech industry may think it has stringent deadlines, but it’s rare that they can’t be pushed. In Formula 1 elite racing, however, it’s just five days — and an ocean or two — between races. If both cars were to crash, new parts likely need to be manufactured, tested and shipped in that time.
“They need to stop what they’re doing. It’s really a true test of agility,” Boyagi said. “We talk about being agile and all these things, [while] they need to stop. They need to pivot. They need to understand which parts they need to manufacture, which is, in itself, quite a long process.
“They need to ship those parts to somewhere else in the world. They need to be tested, they need to be fit into a car and that car has to be driving around the track on Friday.”
While tech’s “artificial” deadlines are usually not that big of a deal, this is a situation, he said, when a missed deadline can be catastrophic.
The Essential Humans in the Loop
F1 pilots are the ultimate multitaskers, driving at more than 300 kilometers per hour on the fastest computers on wheels, receiving data on things like tire and fuel usage, as well as providing data as they go. But they are just an integral cog in a much more complex system.
“What you need is 1,000 people using the same infrastructure, the same software, the same methods in order to work together,” Vowles said, pointing to the benefit of the collaboration with Atlassian. “How do you get collaboration between individuals in order to work towards a unified goal and actually see what each other is doing? And how you can associate with each other and work with each other?”
This will involve not only further adopting Atlassian’s suite of products — which 80% of F1 teams have done — but using it to foster cross-team data standardization and unification.
Unsurprisingly AI is at the center of the AWR engineering strategy, including to create highly regulated simulations for the drivers.
One of the ways, however, that Atlassian Williams Racing distinguishes itself from other tech organizations in 2025 is in its betting on a human-centric future. While other tech companies are looking at AI to replace entry-level jobs, AWR has 130 teammates — 12% of its staff — who are recent university graduates, being trained for the long run.
It’s an interesting time to join F1 too. The whole sport is entering a new era next year, as new F1 regulations will require optimization for lighter, more aerodynamic, fully hybrid cars, which will facilitate not only a smaller environmental impact, but more competitive races.
“We’ve had the same regulations for a few years now, and I think that’s why you see all the teams being very close in lap time,” Button said.
“Next year, there’s a big shift in regulations on the car — also on the engine — and that’s when some people can gain a big advantage.”
Driving DevOps
Atlassian has spent more than 20 years helping technology teams collaborate. Its next step goes way beyond the research and development division.
Whether or not you consider an F1 racing team a tech company — it certainly is now — Atlassian is witnessing a trend that all teams across organizations and industries are becoming agile technology teams. This is enhanced with an influx of AI outside the codebase.
Considering this technological shift, companies from AWR all the way to what the company calls “the Fortune 500,000,” Atlassian’s objectives this year are to:
- Connect all teams and all knowledge workers.
- Make AI part of the team and leverage AI teammates in the near future.
- Become the teamwork partner of choice.
Atlassian is “optimistic and bullish on” the future of AI, said Zeynep Inanoglu Ozdemir, chief marketing officer at Atlassian, in the same press briefing. “We think Atlassian can be synonymous with AI-powered teamwork.”
The Atlassian system of work, which she argues will become a trusted blueprint for enterprise executives, is a philosophy grounded in:
- Aligning work to goals.
- Planning and tracking work across all teams.
- Harnessing collective knowledge.
Ozdemir reflected when last year Atlassian and then Williams Racing were discussing a partnership. She had asked:
- How does everyone know what they should be working on?
- How does everyone align toward goals?
The racing team gave the common response that it tends to revisit goals in meetings once a quarter.
“It’s so different from what we live and breathe in Atlassian,” Ozdemir said, “where everyone has weekly goal checks. And it’s frankly at odds with the very short, agile sprints the racing team follows while on tour.”
For Atlassian, sponsoring a racing team does mean staying on top of mind to, as she said, the 51% of CIOs and CTOs that identify as Formula 1 fans. But it also means being a strategic partner that can help that team accelerate collaboration.
“It was the legendary team of its time [that] is going through a transformation at the moment, changing a lot of things about how it runs, who it runs with,” Ozdemir said. “And they are eager, very eager partners to transform. And luckily for us, they think that teamwork, technology and generally their technology stack is going to make a huge difference in whether they get there or not.”
“This is a partnership where we’re able to actually take a team and really contribute in terms of impact in their success in the next five years. And that’s so much more exciting for us than actually our logo being on a car.”
Add to this, Boyagi said, “Their culture is such a great foundation to build on.”
Currently, Atlassian is supporting AWR in establishing and building its systems of work and understanding how teamwork unfolds on the go.
True Engineering Velocity and Agility
Atlassian will also assist with team-level prioritization, Boyagi said, “helping teams making sure that all their work that they’re doing is aligning to their team goals, which also ladder up to the organizational goals.”
The company will also facilitate knowledge sharing across AWR, he added, “making a lot of the really important information that they create internally discoverable to both other teammates and other teams within Williams, but then also discoverable by AI.”
Finally, it’s not just about delivering physical features at high speed. It’s about delivering high-quality software fast and efficiently, too.
FIA, or Fédération Internationale de l’Automobile, which runs Formula 1 “has a cost cap, so every Formula 1 team has a certain amount of money they have to spend on the development of the car, and that also extends to software engineering and software that they buy and all those things,” Boyagi said.
“Things like developer productivity and delivering software as efficiently as possible is really important for them because it could be a competitive advantage.”
When facing such a restricted budget, this also opens up funds to reinvest in other parts of the car.
Of course, speeding up communication on the fly during race day is something Atlassian’s tooling and DevOps experts can help with, too.
Despite the palpable enthusiasm that comes with joining a new engineering team, everyone reminded the audience that it’s very early in this partnership and too soon to access any productivity, efficiency, cost, or speed gains. Only time will tell if Atlassian Williams Racing makes it first across that finish line.
As Atlassian CEO Mike Cannon-Brookes put it in the keynote fireside chat with Button and Vowles, “There’s not a lot of better ways to talk about high-performance teamwork than through the lens of Formula 1. F1 is a sport that operates on the absolute limit of what’s possible, pushing the boundaries of engineering, but also strategy, teamwork and business.
“Every millisecond counts, and success isn’t just about the driver behind the wheel. It’s about the hundreds and hundreds of people working together, from the pit crew to the engineers, the strategy teams and the corporate officers.”
Disclosure: Atlassian paid airfare and lodging for the reporter to attend this conference.
