Spotify changed what music is — atomic, infinite, geo-spread. It did not change what royalty money is — lumpy, lagged, locked to one country. Praan is the small bet that the second one is about to break, and music is the most natural place for it to break first.
A song on Spotify, JioSaavn, YouTube Music or Apple is played in Lagos at 9:14am, Stockholm at 9:14am, and a marriage hall in Madurai at 9:14am. Three plays, three currencies, three jurisdictions, three sets of split-rules. To the listener, instant. To the artist, a quarterly settlement, in one currency, after a long chain of intermediaries.
The mismatch is the opening. The infrastructure to route a fraction of a rupee to four different contributors the moment a song is played, anywhere in the world — that is what Praan is built around. It batches the dust until it's worth sending, holds the splits in code rather than in a contract folder, and lands the payout in the currency the artist actually spends money in.
Streaming consolidated distribution. The next decade will unbundle economics. What replaces Spotify isn't a better Spotify — it is the layer that lets every play, every replay, every share carry its own micro-economic event. The platform stays. The money flow gets rebuilt underneath.
You can already see early shapes of this — Patreon for monthly tipping, Bandcamp for direct sales, super-fan tiers on YouTube. They each touch one corner. Praan is the part nobody has shipped well: continuous, currency-native, contributor-fair, micro-payment routing — and it sits underneath all of them.
These aren't features. They're shapes the market is already trying to take — and each one needs Praan-style plumbing to actually work.
A new album credits eight people — composer, lyricist, lead voice, three featured artists, mixing engineer, the producer's label. Today they get paid quarterly, on a spreadsheet that nobody reads carefully. With Praan, the second the song plays, the dust splits at the source. Every contributor sees their accrual live.
Indian music — Bollywood especially — is built on remix. Today, the original track holder gets a slow, rough quarterly cut from the remix's earnings. With Praan, the split is written into the song itself: every remix play routes a slice to the original holder, the remixer, the producer, in proportions that travel with the audio file forever.
A Tamil playback singer features on a Punjabi single. The Punjabi label pays the Tamil singer's agency, who pays the singer — three months, two currencies, 30% lost to FX and overheads. Praan lands her share in INR, in her wallet, the Friday after the song streams in Toronto. The currency conversion happens at batch time, at retail FX, not at the agency's markup.
A streamed concert today pays the headliner. It does not pay, in any structured way, the lighting designer who spent a month on the show, the sound engineer who mixed in real time, the choreographer, the dancers. With Praan, every paid second of the stream carries a split — and credited crafts that today are paid once-and-never-again become a continuous tail.
A super-fan listens to the same artist 800 times a month. Today, Spotify pays the artist roughly the same as a casual listener. With Praan, the super-fan can route a small extra dust on each replay — not "subscribe to a platform" but "subscribe to this artist." Listening becomes a continuous expression of support, not a once-a-year T-shirt purchase.
Western music is, structurally, a one-or-two-contributor product. Drake makes a Drake song. There is a producer, sometimes a feature. The royalty story is messy but it's a small messy.
Indian music is the opposite. A Bollywood track has a composer (music director), a lyricist, a playback singer (sometimes two), a chorus, an arranger, a backing band, a producer, a label, often a film studio that owns a piece, often a second label for the regional release. The contributor count per song is structurally 5-10x what it is in the US.
That's not a bug. That's a market shape. The more contributors there are per economic event, the more useful continuous, split-aware micro-routing becomes. Indian music is the market where this matters most.
Add to that: the regional explosion. Tamil, Telugu, Punjabi, Bhojpuri, Malayali, Bengali, Marathi. Diaspora listening from Singapore, the Gulf, the UK, North America. Royalty fragments that today never even reach the artist because they're below the threshold of the aggregator's quarterly minimum — those fragments are what Praan was built to catch.
And the entrenched intermediaries — IPRS, the labels, the music aggregators — are not the wedge. The wedge is the artists who already own their masters: indie tier, DSP-native, hungry for every rupee, and old enough to know how badly the existing system pays them.
The reason Praan is bigger than music — and the reason it sits on Artha, an engine designed for the agentic era — is that the same plumbing solves a much larger problem one step out.
A song is a created thing, used at scale by many, with multiple credited contributors who each deserve a small piece of every use. An AI agent is the same shape. Someone built it. Someone trained the model underneath it. Someone wrote the prompt. Someone curated the dataset. Someone runs the GPU it sits on. When the agent is called — by a person or by another agent — each of those contributors deserves a fraction of a fraction of a rupee.
Today, that doesn't happen. Model fees go to OpenAI. Compute fees go to AWS. Everyone else is uncredited. The same problem music had in 2008, before Spotify made plays at least visible.
Praan in music proves the plumbing for: routing fractional payments, batching dust until it's worth sending, splitting per pre-coded rules, landing in the currency the contributor lives in. The same primitive — once it works for a song — is what every AI agent marketplace will need to be fair to the people who built the pieces.
So when you read the music shapes above, the secret framing is: that is what creator economics looks like in 2031. Music is just where it gets proven first because the listeners already know what a "play" is.
Trust. Does the dust actually arrive? Artists have been promised better royalty deals before. The first thing Praan has to do is show, in a phone they can hold, the dust accruing in real time. That's why the mockup is mobile-first, transparent, and shows the batch about to clear.
Settlement economics. Sending ₹0.0038 over a banking rail is uneconomical — fees would eat the payment whole. That's why Praan batches at the edge: dust accrues until the artist's threshold is hit (e.g. $5 / ₹400), and then a single transaction settles. Continuous accrual, batched payout.
Identity. Proving "this wallet belongs to Arijit Singh, not someone claiming to be him." Solvable, but it's a real piece of work — and labels will try to insert themselves here as the identity custodians.
Tax & FX. Micro-payments × dozens of jurisdictions × four currencies. The boring part. Artha handles it because the whole engine was built around exactly this.
Adoption. Labels and aggregators are the entrenched intermediaries. Going through them is slow. Going around them — direct to artists who already own their masters — is faster but smaller per artist. The wedge is the latter, then the former.
A bet: Praan's first 1,000 artists are not Sony / T-Series / Saregama back-catalogs. They are the independent, DSP-native, master-owning tier. Anuv Jain. Prateek Kuhad. Ritviz. Hanumankind. And — more importantly — the 5,000 artists one tier below them, who today earn ₹3,000-30,000 per month in royalties and feel every rupee of friction in the existing system.
Those artists are reachable through: indie playlists, indie management firms (OML, Universal Music India's indie arm), the YouTube/JioSaavn artist-tools onboarding. They are technologically literate, financially squeezed, and proud of their craft. They are the right first audience.
The labels follow. Once the indie tier is loud about how much better their royalty experience is, the back-catalog opens up as a second wave — by which time Praan has a real network effect.
Praan is shipping in the open. The thinking above is v1 — it will be wrong in places, and we'd rather find out from people who live in this market than from people who don't. Five specific conversations we're actively looking for:
A stream in Lagos. A spin in Toronto. Each play is a grain of dust drifting home — converging, in real time, on the creator in Mumbai.
If you build music, build agents, or think we're wrong — we want to hear it. There is no pitch deck. There is one engineer who reads every email.