The music was always here. Now it is eternal.
Musiki is the world's first AI music platform built on authenticated Vhavenda Tshikona recordings — community-owned, governed by the people who created it, and built as the infrastructure every platform seeking authentic African sound will need to license.
The world's first AI music platform built on authenticated Vhavenda Tshikona. Community-owned. Community-governed. Built as the infrastructure every platform seeking authentic African sound will need.
Three original compositions. Tuned from the exact frequencies of authentic nanga pipes. Owned by a Muvenda woman. Governed by the community that created them.
@MusikiAI · Full 22-Minute Authentic Tshikona Session
Authentic UNESCO-listed Tshikona available for sync licensing. Produced under full community governance and chain-of-custody documentation.
info@musikiai.orgEach performer plays a single bamboo nanga pipe — one note, one person. The melody only emerges when all players interlock their parts in a twelve-pulse cycle. Tshikona cannot be performed alone. It only exists in community — and that is not metaphor. It is acoustic fact.
Tshikona operates on a seven-tone scale derived from hand-crafted bamboo pipes, each tuned individually by ear. No two ensembles are identical. Every performance carries its own set of frequencies. This is what makes it irreplaceable — and what every existing AI model gets wrong.
Performed at the installation of chiefs and kings. Called to summon rain. Played when a community gathers for the most serious occasions of collective life. This is not folk music. It is constitutional music — belonging to the people who made it, on their terms, for all time.
Every AI music platform in the world was trained on the same data — Western catalogues assembled without asking the communities who built the world's most complex musical traditions.
Musiki was built for Tshikona first. Tshikona is not the destination — it is the proof. Once we demonstrate that an AI model can be built on a complex African musical tradition, owned by the community and profitable for its originators, we open the door for every tradition that has been left outside the conversation.
If your tradition is on this list, this is not a curiosity listing. It is an invitation.
You bring the sound. We build the infrastructure. The community retains full governance and revenue rights. The fire stays yours.
Start the Conversation"I am not an outsider who discovered Tshikona through research. I am a Muvenda woman. I grew up at the royal house where this music was always played. That is not a credential. That is a responsibility."
Tshinondiwa Matsila is the founder and CEO of Musiki. She was born in Makwarela, Vhembe, and raised at Musanda Ngwenani ya ha Themeli — the royal house of the Vhavenda people. She heard Tshikona the way it was meant to be heard: in ceremony, in community, played by the people who belong to it.
Fifteen years in data architecture and business intelligence — across FNB, Standard Bank, the National Institute for Communicable Diseases, and the University of Venda — gave her the technical foundation to build what Musiki required. She applied it to the thing that mattered most.
In 2020, she co-authored the Masakhane NLP research paper, supported by UNESCO, IDRC, and GIZ, building machine translation systems across more than thirty African languages. That paper was awarded the inaugural Wikimedia Foundation Research Award of the Year in 2021, presented personally by Wikipedia's co-founder Jimmy Wales.
She built Musiki without a team, without external funding, without permission. Because some things cannot wait.
The Vhavenda are the people of the Soutpansberg mountains in Limpopo Province, South Africa. Their culture is among the richest and most distinct in sub-Saharan Africa — known for their pottery, the sacred Lake Fundudzi, and above all, their music. Tshikona is not simply the national music of the Vhavenda. It is the sound of their collective identity — the one thing that only exists when everyone comes together.
Whether your interest is investment, licensing, or bringing your own tradition into the model — there is a seat at this table.
Musiki is a commercially structured, community-governed AI platform with a live product, documented IP architecture, registered company, BEE compliance, and a 5-year loan repayment model. Active applications to NEF iMbewu, Mozilla Technology Fund, National Arts Council, Ford Foundation, Mellon Foundation, Google.org Africa, and DSI IKS. First-mover position in African cultural AI.
Request the brief →Authentic Vhavenda Tshikona — UNESCO Intangible Cultural Heritage since 2005 — available for sync licensing. Three original compositions are production-ready. Full chain-of-custody documentation and community consent records included. For documentary makers, game studios, advertising agencies, and streaming platforms seeking genuine African heritage sound.
Licensing inquiries →If your community holds a musical tradition that belongs in the age of AI — governed on your terms, owned by your people — Musiki is building the infrastructure for exactly that. You bring the sound. We build the technical and legal layer. The community retains full cultural ownership, governance rights, and commercial participation. The fire stays yours.
Start the conversation →There is a word in Tshivenḓa — zwidade — for the stories told around fires that are going cold. Not the fires of conflict or history lessons. The fires of memory. The ones that hold the sound of a grandmother's voice, the rhythm of a ceremony, the exact weight of a nanga pipe in the hands of the last man who knows how to tune it by ear.
These fires are going cold. Not because no one cares — because no one recorded them. Because no archive holds them. Because the world was not paying attention. Musiki is paying attention now. This is where the fire is kept burning, one story at a time.
He does not use a tuner. He holds the pipe to his lips and listens to the note against the sound inside his own chest — a frequency he has carried since childhood, passed to him by a man who learned it the same way, from a man before him, going back further than anyone living can say. When he is gone, that knowledge leaves with him unless someone is listening right now.
The morning after a chief is installed, Tshikona is played from sunrise. Twelve performers. Twelve pipes. Twelve notes that only make music when they interlock. No one person can carry the melody. The sound itself refuses to exist unless everyone shows up.
In 1956, ethnomusicologist John Blacking arrived in Vhembe with a field recorder and documented Tshikona at a depth no one had attempted before. Those recordings sat in archives for decades. Not hidden — just unheard. The world was not listening. Musiki is listening now.
A tradition does not end when the last performer dies. It ends when no one thought to record them. Musiki began with one question: what would it mean to build the thing that ensures no Vhavenda musician's knowledge disappears without a trace? That question became a platform.
Technical depth without compromise. Community governance without exception.
The VMD-1 is not a generic African music dataset. It is a precision-structured archive of Vhavenda Tshikona recordings, built from authenticated source material with documented chain of custody, community consent, and acoustic analysis at the frequency level of individual nanga pipes.
When an existing AI model encounters Tshikona's twelve-pulse hocketing structure, it classifies it as an error. The polyrhythm is too complex. The scale is too foreign. The model has never been taught that what it is hearing is not a mistake — it is the most mathematically sophisticated communal musical form in southern Africa. We are teaching the model to hear it correctly.
The Masakhane NLP paper — co-authored by Musiki founder Tshinondiwa Matsila — built machine translation systems across more than 30 African languages with support from UNESCO, IDRC, and GIZ. Presented by Wikipedia co-founder Jimmy Wales. The same principles of community-rooted, equitable AI are foundational to Musiki.