How the YRI Fellowship is fueling breakthroughs in artificial intelligence and cloud reliability.
Modern life runs on massive cloud systems. From entertainment platforms and ride-sharing apps to healthcare records and global finance, the world relies on data centers to stay online 24/7. A few seconds of downtime can lead to millions in losses, frustrated users, and in some cases, serious risks to public safety. But what happens when these systems fail? Detecting early warning signs before a crash occurs is one of the biggest challenges in cloud engineering.
This is where anomaly detection comes in. By spotting unusual behavior in real-time data streams, anomaly detection can mean the difference between seamless service and costly outages. For tech giants, banks, and hospitals alike, the ability to predict and prevent system failures is mission-critical. Through the YRI Fellowship, student researcher Aditya Singla is tackling this challenge head-on — and setting a new benchmark in the process.
A rising junior at Cupertino High School, Aditya joined the Fellowship with a passion for computer science, machine learning, and systems software. With mentorship and resources provided by YRI, he took on a problem that even seasoned engineers struggle with: how to improve the reliability of large-scale cloud infrastructure using cutting-edge AI.
Under the program, Aditya developed a multi-step deep learning pipeline that significantly advances the state of anomaly detection. Instead of relying on traditional methods that often produce false alarms or miss subtle warning signals, his system layers multiple techniques to increase accuracy and robustness. The model’s performance speaks for itself.
Using a comprehensive dataset from IBM’s cloud telemetry, Aditya’s framework correctly identified 16 out of 25 anomaly windows — a major leap from the 6 out of 25 predicted in prior research on the same dataset. This threefold improvement is more than a statistical gain; it represents a practical step toward preventing real-world cloud failures. His layered neural network incorporated feature scoring, dimensionality reduction, and iterative suspicion ranking to balance true positives and false positives, striking the delicate trade-off that has long plagued anomaly detection systems.
“I wanted to create something that makes real-world systems more reliable,” Aditya explained. “Through the YRI Fellowship, I was able to design and optimize a framework that helps operational teams catch issues before they escalate, ultimately saving time and resources.” His words highlight the balance between technical sophistication and practical impact — a balance that many industry-grade systems aim to achieve.
The project showcases how the YRI Fellowship empowers students to merge advanced techniques with real-world challenges. Aditya’s framework incorporates tools like epoch optimization and two-stage neural networks, typically seen in graduate-level research, but applies them to the pressing demands of cloud reliability. By automating the early stages of anomaly detection, his model could free DevOps teams from repetitive monitoring tasks, enabling them to focus on the most urgent and high-priority issues.
Looking ahead, the implications of his research are significant. In the near future, Aditya’s approach could scale into automated backend systems that continuously monitor reliability, trigger tests, and self-correct issues before they impact users. For companies operating global platforms, this could mean not just fewer outages, but improved uptime for millions of users worldwide.
The impact of such innovation goes far beyond convenience. In sectors like finance, healthcare, or emergency response, improved cloud reliability can save lives and protect critical infrastructure. The fact that this work is being pioneered by a high school student underscores the transformative power of the YRI Fellowship model.
Aditya’s innovation highlights the Fellowship’s mission: enabling ambitious students to take on world-class challenges and deliver solutions with global impact. Through structured mentorship, access to datasets, and guidance from PhD-level experts, YRI transforms high schoolers into researchers capable of operating at the cutting edge. For Aditya, this translated into not just building a neural network, but into setting a new performance standard in anomaly detection research.
The Fellowship’s approach has already produced results across disciplines — from medicine to climate science to engineering. Aditya’s project stands as another proof point that when young researchers are treated as real scientists, they rise to the occasion. Many YRI Fellows go on to publish in peer-reviewed journals, present at international conferences, and collaborate with academic and industry leaders, even before completing high school.
With leaders like Aditya Singla emerging from its cohorts, the YRI Fellowship continues to solidify its reputation as the fastest-growing research fellowship in the world — one that is not only shaping careers, but also influencing the future of technology itself.
As cloud systems expand and society grows ever more dependent on digital infrastructure, innovations like Aditya’s will be essential. And thanks to programs like YRI, the world won’t have to wait for graduate students or seasoned professionals to push the boundaries. The next generation of pioneers is already here — and they’re still in high school.
