Ballerin-AI: Former Macedonian ballet dancer uses tech to transform performance training in Georgia

Former professional ballet dancer Milka Trajkova is now using artificial intelligence to transform dance training and performance analysis at Georgia Tech’s Expressive Machinery Lab. Her research bridges the gap between artistic movement and data science, creating AI tools that help dancers perfect their technique while making dance training more accessible and precise.

What you should know: Trajkova’s journey from ballet dancer to AI researcher began when a toe injury ended her performing career at age 20 with the Macedonian Opera and Ballet.

• She discovered human-computer interaction (HCI) while studying information systems management, which she describes as “a fusion of design, psychology, and technology” that focuses on “understanding people first, then shaping technology around them.”
• Her undergraduate thesis explored using AI to assess ballet movements, specifically the plié, after finding research on Google Scholar about “ballet technology.”

How it works: Trajkova developed aiDance, a dashboard that converts ballet movements into actionable data for dancers and instructors.

• The system uses machine learning to analyze video footage and identify specific errors in technique, providing visual feedback that supplements traditional mirror-based training.
• During the COVID-19 pandemic, she trained her models using Zoom and Microsoft Teams recordings from ballet classes, extracting meaningful features from ballet motion.
• “When you walk into a ballet class, the highest form of technology dancers use daily is a mirror,” Trajkova said. “I wanted to change that.”

The big picture: At Georgia Tech, Trajkova has expanded beyond ballet to work on LuminAI, an improvisational dance partner that uses AI to transform human movement into new movement.

• This National Science Foundation-funded project culminated in both long-term dance class studies and public performances of AI-driven improvisation.
• Her current research encompasses a broader spectrum of artistic movement, from rhythmic gymnastics to ballet, working with 12 international experts across disciplines including robotics, biomechanics, and sports science.

Why this matters: The applications extend far beyond dance training into healthcare and rehabilitation.

• Dance therapy powered by AI could help with Parkinson’s disease rehabilitation by making physical therapy more engaging and artistic.
• “Most physical rehabilitation has been focused on function, like picking up a cup,” Trajkova said. “We’re asking how we can reframe rehabilitation to make it feel more artistic and enjoyable for people — because the more engaging it is, the more it can support neuroplasticity and lasting recovery.”

What they’re saying: Trajkova emphasizes that her technology aims to supplement, not replace, human instruction.

• “People like to create a kind of mystery around the ballet world, when, at the end of the day, dancers are athletes who need to be well-equipped to perform any role,” she said.
• Her background as a former professional dancer helps overcome skepticism: “I wasn’t a researcher who merely had seen one production of the Nutcracker; I had been in those same satin pointe shoes.”