KAI — Founder Brief

What we're building, why we're building it this way, and why this is the best possible outcome for every person it touches.

The Problem No One Has Solved

For 70 years, the music industry has worked like this:

1. An artist makes great music.
2. Nothing happens.
3. Unless they know someone, pay someone, or get impossibly lucky — a manager, a label scout, a booking agent — nobody with decision-making power ever sees their work.

The internet was supposed to fix this. It didn't. It gave artists the ability to upload — but uploading isn't the same as being seen. There are now 100,000+ tracks uploaded to Spotify every single day. The noise is deafening. The gatekeepers didn't disappear — they just got buried under the avalanche too.

Meanwhile, on the other side of the table:

• Labels are spending millions on human A&R scouts who are still mostly wrong. They're sifting through the same public data everyone else has. By the time Chartmetric flags a rising artist, every competitor has already seen the same chart.
• Venues are booking based on follower counts and gut feel, not actual momentum or audience fit.
• Sync supervisors, brand partners, playlist editors — they all need the same thing: a structured, professional summary of who this artist is, what they sound like, and why they matter right now. Nobody is giving them that in a format they can actually use.

The result: two groups who need each other desperately, separated by a format translation problem. Artists speak in songs, stories, and social posts. Labels speak in pitch sheets, audience demographics, and revenue projections. There's no bridge.

Managers were supposed to be that bridge. But managers are expensive, exclusive, and capacity-constrained. Most working musicians will never have one. And the ones who do still spend weeks manually assembling the paperwork labels need to make a decision.

The Insight Everyone Else Missed

Every music-tech company that's tried to solve this has attacked the label side of the problem: "Here's a better dashboard to find rising artists." Chartmetric, Viberate, Soundcharts, LabelIntel — they all scrape public streaming data, compute a momentum score, and sell the spreadsheet to A&R teams.

They're solving the wrong problem.

The issue isn't that labels can't find artists. The issue is that the data they're finding is:

1. Public — everyone has it, so it's not a competitive advantage
2. Stale — by the time streaming numbers spike, the artist is already being contacted by 10 competitors
3. Unstructured — a Spotify listener count tells you nothing about the artist's story, their goals, their availability, their team, or their readiness to sign

The real asset isn't the data that exists on Spotify. It's the data that exists inside the artist's head — their unreleased work, their creative direction, their tour availability, their financial situation, their dreams. No scraper can get that. No dashboard shows it.

Our insight: the only way to get that data is to give the artist a reason to share it.

What KAI Actually Is

KAI is an AI-powered platform that sits between artists and the music industry. It does one thing better than anything that exists:

It translates.

On one side, an artist talks to an AI manager in plain conversation — about their music, their goals, their story, their next release. They don't fill out forms. They don't navigate dashboards. They just talk.

On the other side, a label executive, venue booker, or sync supervisor sees that same information transformed into the exact professional format they need to make a decision — a structured artist card, a release pitch sheet, a justified recommendation for why this artist fits their brief.

The AI brain in the middle is the product. It's not a feature of a dashboard. It IS the platform.

How It Actually Feels to Use

This is important — because the way KAI feels is as much the product as what it does.

For the Artist: Your AI Manager

You don't land on a dashboard with 47 tabs and a learning curve. You land on a conversation.

The AI greets you like a manager who already knows your story. It asks what you've been working on. You tell it — in your own words, in whatever order makes sense to you.

Behind the scenes, every sentence you say is being translated into structured data: your genres, your influences, your release timeline, your social handles, your collaboration history. But you never see that machinery. You just talk.

The interface spawns what you need, when you need it. Mention a new release? The AI opens a release intake flow. Ask about your royalties? A financial card appears. Wonder how you compare to peers? A momentum view slides in. Nothing is pre-loaded. Nothing is cluttering the screen waiting to be clicked. The experience is fluid, minimal, and alive — like texting a brilliant friend who happens to know the music industry inside out.

When you're not talking, you see one number: your momentum score (0–100). It tells you if you're growing, coasting, or falling behind — relative to artists at your same level, not relative to Drake. Below it: a few daily prompts ("Log today's session", "Post your latest content", "Send a co-sign to someone you respect"). That's it. No overwhelm. No 47-tab dashboard. Just clarity about what matters right now.

The design is incredibly beautiful. Minimal, spacious, with the kind of typography and whitespace that makes you feel like the platform takes you seriously. Not a SaaS tool. Not a social media clone. Something that feels like it was designed for artists, by people who understand that aesthetic matters.

For the Label: Intelligence That Reasons

A label exec doesn't browse profiles. They describe what they need:

"Find me a bilingual rapper under 25 with momentum in Southeast Asia, growing fast, available for a 12-city club tour."

The AI returns a ranked shortlist with reasons:

"RAWLA LITTLE CITY — 18, Papuan-Japanese, Kyoto-based. 10M YouTube plays, +87% MoM Spotify growth. Already has a 12-city Indonesia tour lined up with Willows Media Group. Bilingual Japanese/English. Momentum score: 74 (rising). Estimated $1,774 in unclaimed royalties suggests no current publisher — high signing leverage."

Click through to the artist, and the label sees a live AI-generated card — not a static profile, but a real-time summary regenerated every time someone opens it, always reflecting the artist's latest activity. It reads like a junior A&R analyst wrote it, except it updates itself.

Open the release pitch sheet for the artist's latest single, and the label sees an 80-field professional document — the exact format they'd hand to Spotify editorial, a sync supervisor, or a PR agency. Marketing blurbs in three lengths. Playlist-specific pitching rationale. Campaign hooks. Timeline. All generated from the artist's own words during their conversational intake.

This is the moment the demo wins. The label stops thinking "clever tool" and starts thinking "I would use this on Monday."

Why This Direction — and Not the 10 Other Directions We Could Go

Why not just build a better Chartmetric?

Because momentum scoring is a commodity. Five companies already do it. If we lead with "we score rising artists," investors say "you're another Chartmetric" and the meeting is over in 90 seconds.

The score is a read model on top of the real asset — the AI translation brain and the private data it generates. We have the score too (and it's better, because it includes private signals those companies can't access). But the score is the least interesting part of what we're building.

Why not just build for labels?

Because label-only tools have no moat. If your data comes from public APIs, any well-funded competitor can replicate your dataset in 6 months. Chartmetric's entire business is one Spotify API change away from crisis.

Our data comes from artists using the platform every day — logging sessions, chatting with their AI manager, posting content, sending kudos to peers, planning releases. That's a network effect. The more artists use KAI, the richer the inventory labels see. The richer the inventory, the more labels pay. The more labels on the platform, the more artists join because that's where the opportunities are.

You can't replicate a network effect by scraping an API.

Why not just build for artists?

Because artist tools with no industry connection are glorified journals. Bandcamp, Amuse, even DistroKid — they help artists upload, but they don't help artists get seen by the people who make careers. An artist tool needs to connect to the other side to have gravity.

Why a chatbot-first interface instead of a traditional dashboard?

Three reasons:

1. Artists hate forms. Every platform that asks musicians to fill out 40 fields gets 3 fields completed and abandonment. A conversation gets everything, because humans naturally share information when asked the right questions in the right order. We get 10x more data per session than any form-based onboarding.

1. Context-aware surfaces beat static layouts. A dashboard shows you everything whether you need it or not. A conversational interface shows you exactly what matters right now. Working on a release? You see release tools. Wondering about money? You see the royalty card. Just chatting? You see your momentum and a few prompts. The AI decides what to surface based on what you're actually doing — like a great manager who only brings up the right topic at the right time.

1. It demonstrates the AI brain immediately. The first thing a user experiences is the AI working — not a loading screen, not a signup form, not a tutorial. Within 30 seconds of signing up, the artist is having a conversation with an AI that understands music. That's not just UX. That's the product thesis made visceral in the first minute.

Why royalty recovery?

Because it's the emotional hook no one else has.

Every independent artist with streaming activity has unclaimed money sitting in collection societies, Content-ID pools, and mechanical royalty accounts. Usually $500–$10,000+. They don't know it exists because the system is deliberately opaque.

When an artist opens their KAI dashboard and sees "You may be owed ~$1,774 in unclaimed royalties" — with a plain-language AI explanation of why and where — that's the moment they tell every musician they know about the platform.

It's not a gimmick. The formula is conservative (under-estimates, never over), based on real industry coefficients (WIPO, IMPALA data), and the language is always "estimated" and "plausibly owed." It survives scrutiny.

For labels, the royalty estimate is a discovery signal too: an artist with high momentum AND high unclaimed royalties = unmanaged talent ready to be signed. It's a buying indicator that no other platform surfaces.

And it's a second revenue rail: when we build the real recovery wizard (post-funding, via partnership with a royalty recovery agency), we take 15% of recovered funds. That's pure margin on money the artist didn't know existed.

Why This Is the Best Outcome for Everyone

For artists:

• Access without signing. Universal, Sony, the 1,000-cap rooms, the sync supervisors — they're all on the platform, browsing. The artist doesn't need a manager, a contract, or a connection. They just need to show up and do their work. The AI handles the translation.
• Money they didn't know about. The royalty discovery card alone is worth the signup.
• A manager that never sleeps. The AI keeps them accountable, helps them improve, and structures their career — the things a human manager does, but available to everyone, not just the top 1%.
• Community. Peer kudos, collaboration requests, co-signs. A social graph of people who are actually grinding, not clout-chasing.

For labels:

• Signal 30–90 days earlier than any competitor. The private data from artist-side daily activity is something Chartmetric will never have. First-mover advantage on signing.
• AI that reasons, not just filters. Describe what you want in plain English. Get a justified shortlist back. No more scrolling through spreadsheets hoping something catches your eye.
• Professional-grade artifacts without the work. The release pitch sheet is the format A&R teams already use — but generated automatically from the artist's own words. Hours of manual formatting eliminated.

For the industry:

• Fairer discovery. The momentum formula is designed to be ungameable (tanh-bounded, consistency-weighted, peer-validated). Talent rises on merit, not follower count.
• Less money lost in the system. Billions in royalties go unclaimed every year because the collection system is opaque. KAI surfaces that money and helps artists claim it. That's not disruption — it's correction.
• The manager bottleneck dissolves. The most important relationship in an artist's career shouldn't be gatekept by geography, connections, and luck. AI can do 80% of what a manager does at the translation layer. The remaining 20% (taste, relationships, negotiation) still belongs to humans — but now those humans have better tools too.

For us:

• Three revenue rails, not one. SaaS subscriptions from labels + royalty recovery commission from artists + promoted placement on the matchmaker feed (Phase 2). Each rail is independently viable. Together they compound.
• True network effect = defensible moat. Artist usage -> data -> label value -> more artists. Once this flywheel spins, a competitor can't catch up by raising money — they need to build the network from scratch.
• Every feature we add increases the value of every other feature. More artist activity -> better matchmaker results -> better pitch sheets -> more label trust -> more artist signups. This is the compounding property that makes platforms worth 100x more than tools.

The Three Demo Wow Moments

These are the moments the demo is built around. Each one works for both investors AND label buyers in the same room.

1. The Onboarding Chat (Artist Side)

An artist signs up and immediately starts talking to their AI manager. No forms, no tutorial. Within 5 turns of natural conversation, a fully populated professional profile exists in the database. The artist experienced a chat. The label sees structured data. That's the translation brain working.

2. The Live AI Artist Card (Label Side)

A label opens an artist's profile. Instead of a static page, they see a live-generated summary — elevator pitch, momentum analysis, proof-of-work timeline, endorsements, royalty estimate — regenerated fresh every time. It reads like a human wrote it. It updates itself. That's the intelligence layer working.

3. The Release Pitch Sheet (Label Side)

A label opens an artist's latest release. They see an 80-field professional document — the exact artifact A&R teams hand to Spotify editorial and PR agencies. Marketing copy in three lengths. Playlist-specific reasoning. Campaign hooks. Phase-by-phase timeline. All generated from the artist's own words. That's the format translation working.

Each moment proves the same thesis from a different angle: the AI brain in the middle is real, it works, and it produces professional-grade output that the industry would actually use.

The Momentum Score: How Artists Are Ranked

A single 0–100 number, recomputed daily. Designed to be ungameable.

Formula:

momentum = 100 * tanh(
    0.40 * audience_growth         // Month-over-month listener/follower delta
  + 0.25 * output_signal           // Releases + content posts (log-scaled)
  + 0.15 * engagement_signal       // Saves > shares > likes > views (weighted)
  + 0.10 * proof_of_work_signal    // Sessions logged, quests completed
  + 0.10 * social_co_sign_signal   // Peer endorsements (weighted by endorser's momentum)
) * consistency_multiplier

Why tanh? It caps the return on any single input. A bot buying 100K followers gets almost the same score boost as someone genuinely growing from 5K to 10K. You can't game it by inflating one metric.

Consistency multiplier: Rewards showing up regularly. An artist who works 4 days/week beats one who dumps everything in a Sunday night blitz, even if the raw output is identical.

Bucketed leaderboards (Strava model): Artists compete in three buckets simultaneously — genre, stage (under 1K, 1K–10K, 10K–100K), and trajectory (rising, steady, comeback). Each shows the 5–10 closest peers. Creates competitive motivation without crushing newcomers against superstars.

What's Already Built (as of April 10, 2026)

The demo is on a 7-day build timeline. Everything above "Next" is working code, deployed, tested.

The Demo Artist: RAWLA LITTLE CITY

This is not a fictional persona. RAWLA is a real 18-year-old Papuan-Japanese rapper from Kyoto, Japan.

• Bilingual flow in Japanese and English — blends trap, hip-hop, and R&B
• Viral moment: "Motto Ue!!" collab with Indonesian rapper Paul Shady
• Milestone: First-ever one-man live show at Zepp — Japan's most prestigious concert venue
• Current: Partnered with Willows Media Group (SEA's fastest-growing music company), headlining a 12-city Indonesia club tour in 2026
• Platform numbers: 10M YouTube plays, 28K Spotify monthly listeners, 13K Instagram followers

RAWLA is the kind of artist KAI is built for — talented, grinding, international, and exactly the type that falls through the cracks of the traditional A&R pipeline. In the demo, his data drives every wow moment: the onboarding chat builds his profile, his dashboard shows the $1,774 royalty estimate, and his release "My Dream No Way" generates the pitch sheet that makes labels say "I'd use this on Monday."

Where We Go After the Demo

Phase 1 (Post-Funding):

• Real distribution submission (Spotify/Apple Music API integration)
• Real royalty claims wizard (recovery agency partnership)
• Creative Studio — AI-generated cover art, lyric videos, social cuts
• Production lifecycle tools

Phase 2 (Growth):

• Live performance & touring module
• Direct-to-fan ecosystem
• Vendor & partner marketplace (studios, engineers, distributors as third-party services)
• Ads campaign platform (promoted matchmaker placement)

The "Autonomous A&R" Future: There's a version of this where the AI doesn't just help labels find artists — it identifies breakout trajectories before anyone else and autonomously surfaces deal briefs. We're ~40% of the way there architecturally (the momentum scoring, the matchmaker reasoning, the structured artist data). The remaining 60% (autonomous outreach, deal generation, cross-platform monitoring) is a Phase 2+ expansion that turns KAI from a SaaS tool into an intelligence layer the entire industry runs on.

We don't need to build that to raise. We need to prove the translation brain works. The demo does that.

One Sentence

KAI is the AI manager that translates every artist's daily work into the exact format the music industry uses to make decisions — and gives both sides a reason to show up every day.

Built by Caleb + team. April 2026.