Inside the New Melody Genie: Kevin Affentranger on AI Songwriting, Finished Songs and What Comes Next
Gary WhittakerCreator Spotlight · Melody Genie
Melody Genie has changed substantially since Jack Righteous completed its first hands-on dry run. Founder Kevin Affentranger says the platform now goes beyond lyric development into a finished-song workflow — and the reason for that expansion came directly from watching what writers did after they finished a lyric.
This interview captures Kevin’s thinking behind the product, the songwriting problem he is trying to solve, the philosophy behind Reference the Craft, Not the Person, and what he says has changed in Melody Genie’s newest release.
JR Editor’s Note — August 22, 2026
This Creator Spotlight is based on Kevin Affentranger’s complete written interview with Jack Righteous. Statements about Melody Genie’s newest capabilities are presented as Kevin’s current product statements. Jack Righteous has not yet completed independent testing of the expanded finished-song workflow. Dry Run #2 begins August 22 and will be developed over approximately one week before JR publishes deeper conclusions.
This matters because the original Jack Righteous coverage was built around a different version of Melody Genie: a songwriting-first tool that helped creators develop lyrics, structure, cadence and direction before moving into a music generator. Kevin’s latest answers describe a product that is trying to keep that songwriting foundation while extending the workflow far enough for writers to hear whether the words actually work as a song.
The interview also explains the thinking behind a direction that has become central to both the JR coverage and Melody Genie’s own product guidance: Reference the Craft, Not the Person. Melody Genie has now published its own article on that framework and credited JackRighteous.com for helping shape how the idea is expressed.
Read Melody Genie’s “Reference the Craft, Not the Person” article →This is the full Creator Spotlight, lightly edited for clarity and flow while preserving Kevin’s meaning and voice.
Why Melody Genie Exists
JR: For someone discovering Melody Genie for the first time, what problem did you originally want to solve?
Kevin: Music is getting faster and more generic at the same time, and the lyrics are where you see it first. Production standards keep rising while the writing gets less and less worked on, because the writing is the slow part and everything around it got instant.
The tools did not help. Most lyric generators, and general-purpose language models too, still return the same competent, weightless verse: correct rhymes, no point of view, nothing you would quote back to anyone. They optimise for something that reads finished rather than something that means anything.
The moment it crystallised for me was watching someone regenerate a track ten times trying to fix one line. They could not fix it, because they had never written it. You cannot edit a song you did not write. So I built for the stage everyone was skipping, and the words are the part people actually remember.
JR: What made you focus on the songwriting stage before the music-generation stage?
Kevin: A polished song and a good song are not the same thing, and people confuse them constantly. Production quality got cheap. Anyone can pull a clean mix out of a model now.
What is still hard is having something to say and saying it in a way that lands. If the writing is weak, better production just gives you a well-recorded weak song. So we start at the writing, and the audio comes after, once there is actually something worth singing.
JR: Who did you have in mind when you built Melody Genie?
Kevin: The new generation. AI musicians, rappers who started this year, producers who write their own hooks, people releasing on their phone without ever booking a studio. That wave is enormous right now and almost nobody is building the writing side for them. Everyone is building generators.
It also covers the beginner with a notes app full of half ideas and no structure, and the writer who is good but has been stuck on the same hook for a week. Different levels, same wall.
And a lot of them are not writing in English. A big part of our roster is Spanish, Turkish, French, Portuguese, and those writers usually get the worst tools, because most of this industry builds for English first and translates later.
JR: What do you want a creator to feel after using Melody Genie on a real song idea?
Kevin: I want them to feel like the song got closer to what was already in their head, not like a machine handed them something. If somebody finishes and says, “that is almost it, I just need to change these two lines,” that is the win.
The feeling I am aiming for is momentum. Most song ideas die because the person got stuck, not because the idea was bad.
What AI Should — and Should Not — Do for a Songwriter
JR: What is one thing users often misunderstand about AI lyric tools?
Kevin: That the first output is the answer. It is a draft. The value is not what comes back, it is that you now have something concrete to react to instead of a blank page.
The other one is thinking the tool decides the style. It does not. You decide. The tool is only ever as specific as the direction you give it, which is why “write me a sad song” gets you exactly the song you would expect.
JR: How do you personally define a good workflow between a creator and AI?
Kevin: The creator keeps the meaning. Always. What the song is about, who it is for, which detail is the honest one. That never gets handed over.
The model is good at the mechanical parts: structure, rhyme options, keeping a cadence consistent, giving you five ways to say a line so you can choose. Anything about your actual life, you write. That is the part nobody else can generate, and it is also the only part anyone will remember.
How the Core Tools Fit Together
JR: How do Lyric Writer, DNA Mixer and Beat Matcher work together rather than as isolated features?
Kevin: They are stages of the same job. Write gets you a draft with real structure. DNA Mixer is for when you know the direction but the voice is not landing. Beat Matcher is for when you already have the instrumental and the words have to sit on it.
Most people only need one of them on any given song. If you find yourself reaching for all three on the same idea, usually the idea is not clear yet, and no tool fixes that.
JR: What do you hope creators learn from the artist-inspired writing references in DNA Mixer?
Kevin: I want people to steal technique, not identity. Look at how an artist builds a hook. How many syllables they put in a bar. Where they leave space. How they use repetition, and when they refuse to.
Those are craft decisions you can learn and then use in your own voice. Nobody remembers a song because it imitated someone well. They remember it because it sounded like a person who meant it.
“Reference the Craft, Not the Person” — From the Product Side
JR: What does “Reference the Craft, Not the Person” mean to you inside Melody Genie?
Kevin: The line for me is between the input and the output. The input is openly a name, and it should be. You are choosing whose craft you want to study, and pretending otherwise would just make the product harder to use. Names, photos, the scene someone comes from, all of that is how a writer finds the reference they actually mean.
What comes out is where the principle lives. We are not handing you their voice, their identity, or a way to pass yourself off as them. We are handing you a set of craft signals: vocabulary range, rhyme behaviour, cadence, how a hook is constructed, where the line breathes, the register the delivery sits in. Techniques, not a costume.
That distinction is the same one every musician has always lived by. Studying how somebody writes has never been the problem. Wearing their name has. So the reference is honest and visible, and the result is meant to sound like you learned something, not like you borrowed somebody.
Why BPM Is Not Enough
JR: What makes fitting lyrics to a beat harder than simply knowing the BPM?
Kevin: BPM tells you the speed and almost nothing else. Two songs at the same tempo can need completely different writing, because one leaves the downbeat open and the other starts on the upbeat.
What actually matters is how many syllables fit in a bar, where the stress lands, where you leave silence, and how the phrase ends. Words that look great on paper fall apart the second you try to say them over drums. Anyone who has recorded knows that feeling of a line you simply cannot fit in the pocket, and that has nothing to do with tempo.
A Beginner’s First Test
JR: If a new user has only one song idea and wants to test whether Melody Genie is useful, what should they do first?
Kevin: Pick an artist whose writing you actually like. Type one honest sentence about what the song is about. Generate.
One sentence, not a paragraph of instructions. Then read the draft and change the lines that are not true for you. The whole loop takes about two minutes, and the free credits cover it plus turning it into a song you can actually play.
JR verification note: The free-plan and finished-song details above are Kevin’s description of the current Melody Genie release. JR is testing the updated workflow in Dry Run #2 before publishing an independent assessment.
JR: What result would make you say, “yes, this is exactly what we built it for”?
Kevin: Two things, and they are connected.
The first is when somebody actually learns something. They come in wanting to write like an artist they love, and after a few sessions they start recognising the machinery: why that hook repeats where it does, how many syllables that flow really carries, why the punchline lands on that word. They stop copying and start using the technique. That is the result I care about most, because it stays with them after they close the app.
The second is more immediate. Somebody walks away with a draft that is genuinely worth recording. Not filler, not a demo to prove the tool works. A verse with two or three lines people will actually repeat back to them. If a writer takes that into a booth and records it, we did our job.
JR: What do you want creators to know about revision?
Kevin: The first version exists to be argued with. It is much easier to fix a line than to invent one from nothing, and that is the entire point of the draft.
The edits that matter are usually small. One specific detail instead of a general statement. One real place name. One thing that actually happened to you. That is what turns a competent draft into your song, and it takes about five minutes.
The Founder Story
JR: What part of building Melody Genie has surprised you the most?
Kevin: That it stopped being only mine.
I started this for myself. I am an artist and a trap producer, and I kept thinking about how good it would be to break down how a specific writer actually works — the vocabulary, the cadence, the way they build a hook — and then have that sitting next to me in the studio while I wrote. Like having that person in the room as a writing partner instead of a poster on the wall. That was the whole idea. I built it because I wanted to use it.
The surprise was watching other people want the same thing. Seeing writers I have never met open it, get unstuck, and finish something. That still gets me. You build a tool for your own problem and then find out the problem was never only yours, and honestly that is the most satisfying part of any of this.
The Feedback That Changed Melody Genie
JR: What user feedback has changed the way you think about the product?
Kevin: People kept finishing a lyric they were proud of and then asking the same question: now how do I hear it?
I had been treating the lyric as the finish line. For the writer it was the halfway point. They wanted to know whether the thing actually worked out loud, and until they heard it they could not tell. That is a real gap, because a lyric that reads well and a lyric that sings well are not always the same lyric.
That feedback is the reason we built the music side. You write it, then you press play and hear it sung in that same style, over our production or over a beat you upload yourself. It changed how I think about where our job ends.
JR: How do you decide which features deserve to be built next?
Kevin: Friction beats requests. If I watch someone get stuck in the same place twice, it goes to the top, even if nobody asked for it.
Feature requests tell you what people can imagine. Friction tells you what is actually costing them. And anything that quietly loses somebody — a screen that will not scroll, a button they cannot reach — jumps ahead of every new feature on the list.
The Biggest Melody Genie Update Yet
JR: What are you most excited to improve next for songwriters?
Kevin: We just shipped the biggest update we have done. Melody Genie does not stop at the lyric any more: you write it and then turn it into a finished song, sung in that same artist style, over our production or over your own uploaded beat. That was the missing half of the workflow and it is live now.
We also raised the credits across the board, on the free plan and on every paid tier. The free plan now covers a couple of lyrics plus a full song, so you can go end to end without paying anything to find out whether this is for you. Making people buy before they had heard a single result was the wrong way round.
Next I want to give writers more control over the performance itself. Not just what a line says, but how it is delivered.
JR testing boundary: Kevin’s description above materially expands the scope of the original JR review. Dry Run #1 tested Melody Genie primarily as a songwriting and revision layer before generation in Suno. Dry Run #2 will test the broader workflow rather than rewriting the historical result.
Who Kevin Wants to Hear From Now
JR: What kind of creator do you most want to hear from right now?
Kevin: People who tried it and did not stay. I already know why the ones who stay are staying. I do not know why the others left, and that is the more useful half.
And specifically writers working in languages other than English. That is where I am most likely to be wrong without anyone telling me.
JR: If someone is still unsure whether Melody Genie is for them, what should they try before deciding?
Kevin: Take a song idea you already have and are stuck on. A real one, not a test. Give it one sentence, pick an artist, generate.
If the draft does not get you unstuck, you lost two minutes and no money. If it does, you will know instantly, because you will start editing instead of regenerating.
Kevin’s Closing Note
“I do not think AI should write your songs. I think it should remove the reasons people stop.”
Most people who want to write music never finish anything, and it is almost never a talent problem. They got stuck on a hook. They could not make the words fit the beat. They opened a blank page one too many times and closed it. Those are specific, boring, solvable walls.
If Melody Genie takes those walls down and leaves the real creative decisions with you, it is doing its job. If it ever starts making those decisions for you, somebody should tell me, because it would mean I built the wrong product.
Rapid Fire
What Jack Righteous Is Testing Next
This interview arrives at the right moment for the JR testing process. The existing Melody Genie coverage represents the first version of the evaluation. Dry Run #2 begins August 22, 2026, with the expanded product treated as a new hands-on test rather than retroactively rewriting what happened in Dry Run #1.
The next phase will examine the new song-generation workflow, the relationship between writing and finished audio, uploaded-beat use, revision behavior, where the artist-reference framework holds up in practice, and where Melody Genie fits alongside other AI music tools. The goal is not a same-day reaction. JR expects the deeper test to develop over roughly a week before publishing broader conclusions.
Continue the Melody Genie Series
Disclosure: Jack Righteous has a partner relationship with Melody Genie. That relationship does not replace hands-on testing. Product capability statements attributed to Kevin in this interview are his descriptions of the current platform; JR’s independent findings from the new release will be published separately.
Create What You Love | Love What You Create.
