What happens when you stop letting streaming apps pick your music?

Correction, 10M+ tracks DB and always expanding 

No fillers, No AI content 

OP, Thanks for the writeup, genuinely interesting to see someone trying to attack the popularity-bias problem head-on rather than just tuning it. A few questions before I try it:

What's actually happening under the hood when you say "no AI"? If it's not ML-based similarity, what is the matching logic — metadata tags, personnel/label/producer overlap, something else? "No AI" is doing a lot of work in the pitch and I can't tell if it means "no black-box model" or "no data-driven method at all."

On the Vault: you ran Aretha twice and got >90% different tracks. How is that a feature rather than noise? If the underlying selection criteria are stable, two runs against the same artist should converge on a similar pool, weighted or reshuffled. A 90% swing sounds more like the tracks are being pulled from a low-precision pool than that there's a coherent notion of "70 hidden Aretha tracks worth hearing." What's actually constraining the selection?

How was the depth-level distinction (Essential/Balanced/Adventurous) validated? Genre-adjacency algorithms are notoriously easy to make *feel* adventurous while just reflecting whatever metadata taxonomy you built on. Curious what the actual basis is — audio features, collaborator graphs, something else.

Also, about the "real person, real discography" claim. How are you filtering out the flood of AI-generated tracks and ghost-artist catalogs that Tidal and Qobuz are increasingly full of? If The Vault surfaces obscure tracks by design, it's pulling from exactly the part of the catalog where synthetic content is least policed. Do you have a verification step, or are you trusting the streaming platforms' own filtering (which is known to be weak)?

Finally, about the claim of "no promoted content." Sonic Oracle depends on continued API access to Tidal and Qobuz. If either platform changes terms, throttles, or decides discovery tools like this cut against their own recommendation/licensing economics, what's the contingency? Roon at least has direct commercial relationships with those services — do you?

Not trying to be difficult — I'd actually try this if the mechanism holds up, since Roon Radio's popularity leanings are a real limitation. Just want to understand what's driving the recommendations before I hand it my library.

I’ll also add I’ve bought the app and really enjoy it.  I appreciate all people who take the time to handcraft their playlists.  In fact, I wish we had the ability for people to post their playlists in one place on this website so we could all share in them.  
 

However, I rarely have the time to make my own playlists.  Sonic Oracle has done a great job of filling that gap for me. 
 

it would be interesting to learn what Sonic Oracle’s top 5 searched for artists might be over a weekly or monthly period.  Maybe even get a representative playlist based on how users are searching for them.  
 

The app is one more tool to help me enjoy the music and I am appreciate of it.  

Good questions. I'll go through them one by one.

No AI: It means no black-box ML model deciding what you hear. The engine is data-driven. It uses structured signals from a proprietary database, but the logic is deterministic and inspectable, not a trained model. "No AI" also refers to the music itself: no AI-generated tracks, no synthetic filler.

The 90% Vault swing: Fair challenge. Two runs pull from a deliberately large qualifying pool, then vary the sampling. So you get a different valid slice each time, not noise. The pool is constrained by relevance filters. The shuffle is intentional so repeat runs stay fresh rather than identical. You're right that a 90% swing sits on the edge between intentional variety and low-precision pool. I'm not going to overclaim coherence here. It's tuned to lean toward variety because the whole point is hearing something different every time. If convergence is something users want, it's a knob I could expose.

Essential/Balanced/Adventurous: Based on catalog-relationship distance, not audio features. How far the recommendation sits from the seed across the signals the engine uses. You're right that adjacency can feel adventurous without being so. Honest answer: it's validated by listening and user feedback, not a published metric. 350+ paying users and growing, and the depth levels are the feature I get the most positive feedback on, especially Adventurous. But no, there's no formal benchmark behind it.

Ghost artists and AI tracks: Legitimate concern, and you're right that deep cuts are exactly where synthetic content hides. I run my own verification process rather than trusting what the platforms filter out. It's active and ongoing, not a one-time claim. Not perfect yet, but it's mine, not theirs.

API dependency: Straight answer. No, I don't have direct commercial deals with Tidal or Qobuz like Roon does. It's a real dependency risk. Contingency is multi-platform support. Already live on Tidal and Qobuz, YouTube Music coming soon. No single platform is a point of failure. And the selection logic is platform-independent. It doesn't rely on their recommendation APIs, only library access and playback.

Give it a try. Pick an artist you know well, set it to Adventurous, and see if the mechanism holds up for you. I think it will.

Alessandro