She may not know her way around a semi-automatic sequential gearbox, but thanks to artificial intelligence (AI), one Formula One fan is leaving the odds in the dust when it comes to predicting the winners of Grand Prix races.
Armed with an algorithm and a spreadsheet, data scientist Mariana Antaya has correctly predicted three out of five winners so far this season.
Ms. Antaya, a product manager with Microsoft in Redmond, Washington, built a machine learning model and fed it a range of data points — performance stats from the 2024 season, qualifying results from 2025, weather forecasts, lap times and other information sourced from the FastF1 API — to fine tune her digital fortune-teller.
“I got into it because I’m a big fan of F1. I grew up watching it, and my uncle was a race car driver. So with my background, I thought it would be cool to build a model to predict the winners,” she said in an interview with Techstrong. Ms. Antaya plans to start predicting the entire podium in the near future — that’s first, second and third place. AI models rely on sophisticated algorithms to digest complex data points for analysis and prediction. Such models are already used in finance, medicine and sports.
“As far as using the information for betting, I wouldn’t recommend it,” Ms. Antaya said. “After all, you can’t predict the future. There’s a lot of uncertainty. There are sports betting sites out there for that — what I put out is more educational.”
With her trademark hair fling at the start of each video, she offers tutorials. She recently posted a segment on how to use GitHub to fine tune an AI model. “And I’m not even kidding, I used to be that girl who would just be playing around with LLMs but never actually built a project or deployed something, and this is your answer to that.”
Ms. Antaya said she’s been surprised by the attention her predictions have received. “A lot of people have started making their own prediction models. They post them on social media and tag me, so that’s awesome,” she said. “They’re building off the base model I created, and it’s fascinating how you can improve the model or change it to your liking.”
The mean absolute error (MAE) for her model is improving. On April 5, prior to the Japanese Grand Prix — where she correctly predicted Max Verstappen would win — she noted, “If you remember from our first race, I think our first MAE was around 3.4 seconds, and now we’re at zero-point-five seconds, which is kind of crazy. That means adding these new features is improving our model.”
Now, she’s crowdsourcing ideas from fellow fans to make her system even sharper, inviting suggestions for new variables that could help her predictive streak keep pace with the ever-unpredictable world of the Formula One Grand Prix circuit. “Crowdsourcing has been going way better than I expected. So many fans are giving me suggestions — recommendations of things to add — so the model should get better week to week,” she said.
People close to F1 racing are also reaching out to her. Engineers already use AI to help make vehicles more aerodynamic, predict the best pit stop windows, select optimal tires for track and weather conditions and refine race strategy.
“Some of the engineers called me, so I’ve been able to learn a lot more in terms of what they’re doing,” she said. “So now I can take what I’ve learned and improve the model based on their model. The next videos will have inclusions from real F1 engineers and what their input has been to the model, which is super awesome.”
The McLaren racing team sits atop both the driver’s and constructor’s standings (prior to the Miami Grand Prix on May 4), with Oscar Piastri currently leading in points followed by his teammate Lando Norris in second place.
McLaren is in a partnership with Dell Technologies, to use data to improve performance. McLaren says it has collected racing data since the 80’s, but with the rise of AI, can now truly put it to work.
“AI has enjoyed a rapid rise to prominence in recent years, exploding across the wider world and developing even faster than people’s understanding of how to use it,” the team stated on its website.
“Unlike many inventions from the last few decades, it came with no user manual but so many possibilities, especially in Formula 1, which shares one major trait in common with AI: it loves data. Our team wasted no time exploring its potential benefits and has gradually introduced it into our everyday lives, to the point that we now no longer know how we’d live without it. Our team loves data, but we’re only human and can only process so much, both due to time constraints and because too much data is simply overwhelming and counterintuitive. But not for AI, which can digest and cut through it, telling us precisely what is useful and what we can disregard. And the more data we feed AI, the smarter it gets. Rather than keeping huge libraries worth of statistics that go largely unutilized, AI can tell us what we need and synthesize it into actionable insights. The time saving is astronomical. And not just time savings producing existing reports, but brand new types of learning are possible that we could have never attempted before AI. So much valuable data would have gone to waste.”
For Ms. Antaya, tech is her work and hobby. She grew up surrounded by it — her parents worked in the industry, and it was heavily emphasized in her education from elementary through high school. She took coding classes and joined the robotics team in high school. Her father gifted her a book on deep learning long before it became the popular catchphrase to describe how AI gets smarter. By the time she applied to college, she was zoned in: Computer science.
Her career has since flourished, and now she may have found a new lane blending her knowledge of AI and machine learning with Formula One. When asked if she would like to get more involved with the F1 organization, she said, “I’d love to. That would be incredible.”
She attended a Formula One race in Barcelona a few years ago and has visited tracks in Abu Dhabi, Monaco and a few other places. Her favorite driver is Lewis Hamilton with Team Ferrari.
“I just need a Lewis win — that would make my season,” she said.
