Contents
MuseCool is rethinking what a music school can look like in a digital era
Music education hasn’t changed much in generations. Children still attend weekly lessons, practise inconsistently at home, and teachers rely largely on instinct rather than measurable data. But now a startup believes AI can fix that gap and bring gamification and feedback loops to one of the most traditional corners of education.
MuseCool is a London-based music education company founded in 2017 that provides personalised music lessons for students of all ages, both in person and online.
The company connects learners with professional, conservatory-trained tutors for instruments such as piano, guitar, violin, and drums, and supports exam preparation, performances, and workshops.
MuseCool combines traditional teaching with AI-supported practice tools to help students progress between lessons, positioning itself as a modern private music school focused on flexibility, high-quality instruction, and making music learning more accessible and engaging.
I spoke to CEO Petru Cotarcea to learn more.
Cotarcea grew up poor in Romania. But because of the Communist legacy, music education there was extremely cheap. “That meant I could learn an instrument, which wouldn’t have been possible in the West,” he shared.
“It was one of the few good things about that system.”
From scholarship student to West End performer Music became Cotarcea’s pathway to opportunities he wouldn’t otherwise have had.
By 14, he was already competing internationally and had earned a scholarship to Chetham’s School of Music in Manchester, one of the UK’s most prestigious boarding schools for young musicians. He later continued his training at the Royal Academy of Music and quickly entered the professional world.
He recalls:
“By 18, I was playing in the West End production of Sweeney Todd and met Stephen Sondheim. I was performing with professionals much older than me, and I realised I’d been lucky — right place, right time. It made me think about what came next.”
A failed peppermint farm — and a pivot
After his first year in the West End, Cotarcea had saved about £10,000, “which felt like a lot of money at the time”.
At 19, he invested in a peppermint farm in Romania. The logic was that he could keep studying in London while this grew in the background. After a year, Cotarcea planned to apply for grants and build an essential oil processing facility.
However, he revealed, “the crop turned out to be toxic. It didn’t contain menthol, but it contained a chemical illegal in the EU. The lab told me I had to burn it. So I literally burned the entire field. That was my first venture.”
This failure convinced him to build something in a field he understood: music education.
Cotarcea now runs one of the largest music schools in London, with operations in the UK and New York. It aims to change how music is taught.
The school’s audience is children aged 5–14.
The problems with traditional music teaching
The fundamental problem is that music education is still very traditional, kids don’t practise enough, and there’s almost no data on what actually works.
“If they don’t practice, they don’t improve. If they don’t improve, parents stop paying for lessons. Fix practice, and you fix the entire system,” asserts Cotarcea. Cotarcea describes music education as deeply traditional.
“Teachers tend to teach the way they themselves were taught — often following methods passed down for generations.
There’s real pride in lineage: my teacher taught me this method. Kids who enjoy their lessons often stay for years. Piano dominates the beginner market, accounting for roughly two-thirds of new students.”
However, music teaching has largely failed to adapt to a more digital-first world and fails to reflect how children today learn.
Why audio AI is harder than text AI
Put simply: audio isn’t discrete like text — it’s a moving, layered signal, and teaching machines to understand it is closer to decoding a performance than reading a sentence. According to Cotarcea, Audio AI is lags behind text and image AI.
“There are very few mature tools for music understanding. Most of what we built is based on research and custom development.
The big challenge is interpreting imperfect human performance. If a child plays slightly out of time or tune, humans recognise it instantly. Machines don’t.”
‘A big challenge is interpreting imperfect human performance. If a child plays a snippet of music, humans can recognise it — and its mistakes — instantly. Machines don’t.’
MuseCool’s approach is different. Teachers press “start” at the beginning of a lesson and “finish” at the end.
“We listen to the lesson, understand what happened musically, and automatically generate practice games and analytics. This empowers teachers to better understand the individual needs of each student and plan accordingly,” shared Cotarcea.
Gamifying music education
MuseCool’s flagship product is The Muse, an AI-powered practice assistant that supports music students between lessons by turning what was taught into guided, motivating home practice. It uses lesson data to generate short, personalised practice sessions that feel more like games than repetitive drills, helping children practise more consistently and with better focus.
Parents receive practice analytics and simple progress updates, while tutors can seamlessly integrate the tool into lessons, keeping practice aligned with what was taught and boosting continuity week to week.
Early software testing revealed a surprising reality — beginner lessons contain less playing than you’d expect. There’s a lot of conversation, encouragement, and attention management, especially with young kids.
“Historically, no one has had large-scale data about what happens inside music lessons. If we scale, we could reshape music education research simply by providing real data,” asserts Cotarcea.
A future built on data and scale
Looking ahead, Cotarcea sees MuseCool evolving on two fronts: a global practice platform and a data-driven marketplace.
“There are two layers,” he explains.
“First, a global platform where tutors can use our platform for free while parents subscribe. The goal is simple: kids practise better, and lessons become more effective.”
The second plan is a teacher marketplace. As tutors adopt the platform worldwide, the school can match families with teachers using real performance data, creating smarter connections than traditional directories.
The company is already testing the model at its large London music school, which serves as a live ecosystem for product development before a public launch in March and a wider international rollout.