
Music can change pace, intensity and mood in a few seconds. Anusha Arora wanted a therapy platform that could respond just as quickly.
The 17-year-old student from Washington state built HARMONI, short for Harmonically Adaptive Real-Time Multimodal Output for Neural Integration. The project uses artificial intelligence to adapt a music-therapy experience using more than one kind of input.
At ISEF 2026, Arora won the $50,000 F. Thomson Leighton and Bonnie Berger Family Prize for STEM Excellence.
Digital health products often deliver the same programme to everyone. HARMONI’s title describes a different aim: adjust the output in real time as the system receives information about the person using it.
That could make an intervention feel more responsive. It also creates new questions. Which signals does the platform use? What counts as improvement? Does the system adapt in a way a therapist can understand and supervise?
Mental-health technology carries a higher burden than an ordinary music app. A polished experience or positive reaction does not prove a clinical benefit. Studies need suitable comparison groups, clear outcome measures and safeguards for people whose condition may worsen.
ISEF gave HARMONI one of its largest 2026 prizes. Arora qualified through the Central Sound Regional Science and Engineering Fair and presented alongside more than 1,700 finalists.
That result shows expert judges found the research exceptional within the competition. It does not turn the platform into approved therapy.
The interesting next chapter is the evidence: whether other researchers can reproduce the findings, whether the system works for different groups and how clinicians judge its place alongside human care.
Arora’s achievement sits at that boundary. She has built an award-winning technology around a difficult health question. The prize is an outcome. Clinical value remains a question for much more testing.
Tradeoffs
- Real-time adaptation may make an experience more responsive, but can make the system harder for a clinician to inspect.
- A large science prize validates the research achievement, not the platform as approved therapy.
Try this
What should HARMONI measure before anyone describes its music as therapy rather than an experience?
