Eighteen-year-old Matteo Paz, a high school student, developed an AI-based model that identified 1.5 million new celestial objects using NASA's retired NEOWISE telescope, a significant achievement that earned him a $250,000 prize.

The algorithm has attracted interest from professional astronomers and could also be applied to analyze trends in stock markets or climate data.

The algorithm detects subtle variations in infrared brightness over time, allowing for the processing of the entire NEOWISE dataset.

Despite NEOWISE being retired, its archive of nearly 200 billion observation rows contained valuable, unexplored data on various astronomical phenomena.

Paz applied his knowledge of AI and advanced mathematics to develop a machine-learning model capable of efficiently analyzing NEOWISE's extensive dataset.

As of 2025, Paz works at Caltech, gaining hands-on experience in data management and analysis for NASA and NSF-supported missions.

Within six weeks, Paz created an AI model that flagged and classified a vast number of potential variable objects in space, demonstrating significant promise.

Paz's groundbreaking work has been recognized in The Astronomical Journal, showcasing the effectiveness of machine learning in analyzing large astronomical datasets.

The refined AI model is set to publish a comprehensive catalog of variable celestial objects detected in NEOWISE data in 2025, with potential applications beyond astronomy.

Paz's experience in the Regeneron Science Talent Search provided valuable insights into mentorship, reinforcing the importance of nurturing young talent.

Kirkpatrick initially suggested examining a small area of the sky for new variable stars, but Paz opted to automate the process using his AI and mathematics skills.

Davy Kirkpatrick, a senior scientist at IPAC, mentored Paz and has a history of guiding high school students in scientific research.