FAQ

WAV, FLAC and MP3. For best results, send the highest quality file you have.

Contact us at kolacko@soundintel.tech.

Anyone who decides what happens to a song before it is released. Labels and A&R teams choosing which track to push and where to spend the marketing budget. Artist managers planning a release. Distributors prioritising large volumes of incoming music. Sync agencies matching tracks to briefs. Publishers evaluating catalog potential. And independent artists who want to know how their song lands before the world hears it.

No. Your music stays yours. Audio files, released or unreleased, are never shared with any third party. Audio files are deleted after the analysis is completed (as a rule within 30 days, or as agreed with you). The only thing we keep is the sound fingerprint of your track, as documentation of the analysis performed and for follow-up analyses you may request. By default it is deleted after 24 months, and you can have it deleted earlier at any time. Neither your audio nor anything derived from it (fingerprint, extracted characteristics, reports or results) is ever shared with third parties or used to train any external or generative AI model. The only exception is SoundIntel’s own internal analysis model, and only where the rights holder has given explicit written consent in a separately concluded agreement.

A set of measured acoustic characteristics of your track, such as tempo, key, energy and spectral profile. It is not the recording itself: your song cannot be reconstructed or played back from it. We keep it as documentation of the analysis and to make follow-up analyses possible. It is never shared and never used to train any external or generative AI model. It may be used to improve SoundIntel’s own internal analysis model only with the rights holder’s explicit written consent in a separately concluded agreement.

By default, the fingerprint is kept for 24 months after the analysis, as documentation of the analysis performed and for any follow-up analyses you request, and then deleted, unless otherwise agreed. You can ask us to delete it earlier at any time by email to kolacko@soundintel.tech, and we will confirm the deletion in writing.

No. A fingerprint contains no audio at all. It holds only technical data about the audio file that matters for the analysis, such as tempo, key, energy and spectral profile. That data describes the track; it is not the track. Your song cannot be reconstructed, played back or recreated from it.

Uploaded audio files are deleted after the analysis has been completed (as a rule within 30 days, or as agreed with you).

Uploaded recordings are kept in encrypted, private cloud storage provided by Vercel Inc., which acts as our processor and may process data outside the EEA (including the USA) under the EU-U.S. Data Privacy Framework and EU Standard Contractual Clauses. For the analysis, they are transferred to SoundIntel’s own internal secured server in Austria (EU). Audio files are deleted automatically after the analysis is completed (as a rule within 30 days, or as previously agreed with you).

No. Uploaded recordings are used only for the agreed SoundIntel analysis. They are never shared with, sold or disclosed to any third party, and never used for marketing. The only outside companies involved are our infrastructure providers (Vercel for hosting and storage, Resend for confirmation emails), which process data strictly on our behalf as processors under GDPR and may not use it for their own purposes.

No. Your audio and any data derived from it (sound fingerprint, extracted characteristics, reports or results) are never used to train external, third-party or generative AI models. The only exception is SoundIntel’s own internal analysis model, and only if the rights holder has given explicit written consent in a separately concluded agreement. Without such an agreement, your track is used solely for your analysis.

Access is restricted to authorized SoundIntel personnel and systems required to perform the agreed analysis.

No. SoundIntel analyses music, it does not generate it. Your music is never shared. Audio files are deleted after the analysis is completed (as a rule within 30 days, or as agreed with you).

You must either own or control the necessary rights to the recordings and compositions, or be authorized by the relevant rights holder(s) to submit them for analysis. If you are submitting on behalf of an artist (as a manager, label or A&R), written permission from the rights holder is sufficient. We may ask for proof. Responsibility for the submitted material remains with the sender.

Yes. The Track Submission Agreement and Privacy Notice are available directly in the submission form before you confirm and upload your files.

Yes. Contact us at kolacko@soundintel.tech.

Send release-ready tracks. The analysis measures the sound itself, so recording quality directly affects the result: a rough mix or unpolished demo will not produce the clear, reliable analysis a finished track will.

Yes. Each version is analysed and billed separately. Contact us for a combined offer.

Yes, no prediction is ever certain. Across our internal tests, SoundIntel’s audience predictions matched the observed outcome in around 80% of cases. In a public, timestamped test before the Eurovision Song Contest 2026, SoundIntel identified 15 of the 20 semi-final qualifiers (75%) from audio alone, with no artist names, countries or betting odds. Full results are on our validation page. The report is a strong input for your decision, not a guarantee.

SoundIntel has conducted internal validation and published preliminary blind-test results. Detailed technical validation materials are available to qualified investors and strategic partners under NDA. Public examples are intended to demonstrate specific capabilities rather than establish universal predictive accuracy.

SoundIntel’s methodology is proprietary intellectual property. We provide sufficient public information to explain the scope, intended use and limitations of our technology, while offering confidential technical due diligence to qualified parties.

Streaming data tells you how a song performed after release, when the budget is already spent. SoundIntel works before release: it simulates how audiences will receive your song, and which audiences, while you can still act on it.

No. SoundIntel is an intelligence tool: it shows you how audiences respond and where your song's strongest audience is. The decisions stay yours.

SoundIntel grows every day. It is strongest in established genres and also recognizes niche styles; how your track was classified and analysed is explained in the report.

Yes. SoundIntel analyses the sound of the recording itself, so it works for music in any language.

The report tells you how audiences receive your song, not how to write it. Each analysis is reviewed by our team before delivery, and where we notice something worth flagging, it is noted in the report.

We are building toward large-volume analysis. Contact us at kolacko@soundintel.tech to discuss your catalog.

Every report includes a sync-fit indication. Scoring against a specific brief (mood, scene, target audience) is available as a custom analysis, contact us at kolacko@soundintel.tech. We are also building Proszenium, a model built on SoundIntel that can watch the video itself and judge even more precisely how well a track fits each scene. More on the Proszenium website.

Not yet. Integration for catalog and distribution partners is on our roadmap; if this matters to you, contact us at kolacko@soundintel.tech and we will keep you in the loop.