Hazel Savage defines herself as a punk rocker by heart. Her love for music led her to start a career in the music industry, where she worked for renowned companies like Universal Music, Shazam, and Pandora. At the same time, she also played guitar for an all-girl punk band in London named Ginkinta.
Her career of over 14 years in the music industry gave Savage the opportunity to leave her home country of England to work in other nations such as Australia and Singapore, where she decided it was time to take the next step. In 2018, she joined a startup incubation program called Entrepreneur First, which tech talents can connect and create new startup projects.
During the program, Savage met Aron Pettersson, a Swedish data scientist with over 17 years of experience in artificial intelligence and data science. The two decided to launch Musiio, a platform that leverages deep learning and AI to assist music companies in curating their music catalogs.
One of the biggest pain points in the music industry is scrutinizing the massive volume of tracks that are being created or released every day, Savage said. For example, 60,000 tracks are ingested by Spotify daily. On average, a new track is uploaded every 1.4 seconds to the Sweden-based platform.
At such speed, record labels, music streaming firms, and music archives have found it almost impossible to tag and adequately categorize all of their music into different playlists. That’s where Musiio comes in. The platform’s AI technology can read and extract various features and patterns of songs by analyzing the instruments in the song, the tempo, the genre, and other elements. According to the company, it only takes Musiio a day to tag 5 million songs, whereas a music curator can only go through 100 tracks in the same amount of time.
Empowering music curators
Musiio says its software allows streaming companies to optimize their tagging capabilities, playlist creation, and search functions of new music. The company also wants to help talent scouts find new artists by providing filtered lists of songs according to their requests.
Currently, the startup counts notable names as clients, including US-based music production marketplace BeatStars, London-listed music rights investor Hipgnosis Songs Fund, and Vietnamese music gaming startup Amanotes, the maker of piano game Magic Tiles 3. Musiio’s main stream of revenue comes from its track tagging and search software, which is available either on a monthly subscription basis or as a one-time purchase.
“Within the first two years, we managed to increase our revenue by over 600%. Currently, we have been able to double our revenue every year, which is a testament to our products,” Savage said.
Although the startup has already reached profitability, Savage is working to increase the percentage of recurring income. “So far, half of our profits are recurring, so I wouldn’t say that we are profitable yet. I would say we are breaking even.”
In July, Musiio reached a USD 10 million valuation after raising an undisclosed amount in a fundraising round with the participation of Francois Arbour, the founder of royalty-free music website Premiumbeat, and Paul Kempe, the owner of management, publishing, and record company Tileyard Group. Currently, Musiio is exploring options to expand its business into the UK and the rest of Europe.
Similar AI music startups have emerged in recent years, including French-based Niland, UK-headquartered Sonalytic, and German startup Cyanite.ai. Niland and Sonalytic were acquired by streaming giant Spotify in 2017 and 2018, respectively. Savage said she is open to Musiio being acquired by a major music service too.
“AI is a very cost-intensive product. If you’re not going to raise any money, then you will only be able to grow slowly. If you have strong opinions regarding fundraising, mergers, and acquisitions, it can be tough to take on growth opportunities as they occur. The best thing I can do is keep an open mind because that’s when you get to grow,” Savage said.