Show HN: Deep Cuts – music discovery via credits instead of algorithms

A new platform called Deep Cuts is reshaping how listeners find music by focusing on credits and human connections rather than traditional algorithmic suggestions. Launched by developer Xenia Dobonyi, the site allows users to explore music starting from a record or mood, following the people involved in its creation to uncover related works, all while explaining the reasoning behind each recommendation.

Deep Cuts diverges from current streaming services where recommendations rely heavily on passive listening habits and present lists without context. Instead, it utilizes detailed credit information sourced from open databases like MusicBrainz and Discogs, linking users to streaming previews primarily through Deezer and Apple Music. This approach aims to deepen listeners’ understanding and appreciation by revealing the creative networks behind songs and albums.

The interface plays short previews and enables users to save favorites in a personalized crate, which also helps refine future suggestions, ensuring no repetition. Visits and use are tracked anonymously, maintaining privacy without requiring accounts or ads, and interactions remain stored only within the user’s browser.

Behind the scenes, the curation process involves artificial intelligence that selects and justifies recommendations based on real credit data. The AI models, including Claude Haiku and a backup open model on Groq, produce explanations that encourage verification through physical or official sources like Discogs, emphasizing transparency.

This system faces limitations due to incomplete credit documentation, especially for newer releases or genres with less thorough metadata. Classic soul, jazz, and rock records tend to be better served, but users are advised to input album titles or descriptive moods to enhance results when single tracks yield sparse outputs.

The platform's philosophy emerges from the belief that musical taste can evolve through guided discovery, rather than purely algorithm-driven repetition. By connecting listeners directly to the artists, producers, and session musicians involved, Deep Cuts fosters an active form of exploration rather than passive consumption.

This credit-centric approach holds potential repercussions for artists’ catalogs and how their work circulates in the digital era. It could elevate overlooked contributors whose roles traditionally go unrecognized in mainstream recommendation systems, thus broadening exposure beyond headline artists.

For touring cycles, understanding the web of creative participants might inspire new collaborations or live performance lineups featuring session players or co-producers highlighted through these explorations. This might also affect how rights holders and managers consider metadata quality and completeness.

From an economic standpoint, genres with rich archival metadata could experience renewed interest, impacting sales, streaming, and licensing opportunities. Conversely, scenes with poor credit coverage risk being underrepresented, underscoring the ongoing importance of comprehensive documentation practices.

Deep Cuts operates independently from the platforms it links to, without official affiliation or sponsorship by Discogs or similar entities, relying on publicly available APIs and data. Its open-source nature invites community involvement and iterative improvement.

The project exemplifies an alternative method to AI-assisted music discovery, prioritizing context and human collaboration over pure algorithmic prediction. It suggests a pathway for listeners who want to move beyond surface-level recommendations and engage more meaningfully with music’s creative ecosystem.

As the platform grows, the quality and breadth of credit data will determine how expansive and precise its suggestions become, potentially influencing industry attitudes toward metadata stewardship and fan engagement strategies.

Future enhancements could include broader integration with music databases and more refined AI models, but the current iteration already distinguishes itself by emphasizing transparency, privacy, and user-driven exploration intertwined with AI assistance.