Italy based company Musicmatch Well known for providing community backed lyrics to major music streaming platforms including spotifyand Apple Music, YouTube Music, Amazon Music, and Tidal. Launching now New podcast platform Which combines AI-generated transcription with community-verified editing.
While there are millions of podcasts and episodes available to listeners, Musixmatch says podcast search is disabled. As a result, it indicates that a lot of great podcasts don’t connect with potential fans. So she is using her experience training AI models through song lyrics and leveraging her expertise in NLP (Natural Language Processing) to improve podcast transcription, search, discovery, and sharing.
Musixmatch’s podcast platform automatically generates versions every day of some of the best podcast episodes across different topics and infographics. It uses the architecture of the underlying NLP model, Umberto, to tag keywords such as places, people, and topics with Wikipedia IDs – alphanumeric identifiers associated with topics on Wikipedia. (for example, this link Denotes a Wikipedia ID related to TechCrunch.)
Because of this approach, people who search for these topics in any language will get accurate results.
The startup explained to TechCrunch that based on these identifiers, it creates a graph of topics called TopicRank that ranks podcasts based on factors such as the number of mentions in an episode or the expertise of the presenters on the topic — improving search results for podcasts when users search for related topics.
Thanks to this classification, people can finally search for any given keyword and find written podcasts that match their query, ranked by relevance. Our search index shows a more detailed and in-depth set of results than any other listening services that rely on standard RSS metadata. and pre-defined types and categories.”
When users search on Musixmatch’s podcast platform, it shows excerpts from transcripts where the searched phrase is mentioned. If they click on the result, the audio will start directly from the timestamp of the passage referring to the phrase. This is very cool when you need to listen for a few minutes of audio while searching for something.
Musixmatch has always relied on its community to make subtle tweaks to lyrics, and now it’s asking those users to do the same with podcasts. The company’s new podcast portal also includes a tool called Podcast Studio, which allows editors and podcast owners to fix AI-generated copies — particularly useful for things like people, brand names, or cultural references.
If there is no script for a particular episode, the owner or a community member can use Podcast Studio to create a specific episode. Musixmatch says that the AI takes approximately five minutes to generate a script for an episode. Regular listeners can also vote for an episode to transcribe so that the community prioritizes it.
It’s important to note that on the Musixmatch platform, AI-generated transcripts will contain tags like “Speaker 1” and “Speaker 2”, while community-edited episodes will have speaker labels – along with a tag Verified.
The company is also making sharing easier by displaying cards that contain text excerpts from podcasts Shareable link. Moreover, it is working on a feature called audiograms, which are small shareable videos that include audio excerpts and scrolling scripts from podcasts.
Musixmatch does not want to keep all this data to itself. Allows podcast owners to export transcripts to their own web feeds and apps. And since these texts are SEO-friendly, she argues, they’ll make it easier for listeners to find them.
Among Musixmatch’s partners that it says use its copy tools are “The Talent Show” from The Financial Times; “Beyond the Ordinary” and “Why I Run” from Red Bull; and an entire Chroa Media production.
While the podcast platform Musixmatch offers features to listeners, it is not trying to be just a podcast player. The startup says its competitors are companies working in audio analytics — including apps that provide transcription services (such as podcastle).
“We believe that voice analysis (artificial intelligence, semantic, etc.) will be essential in the near future, for many different use cases. We are uniquely positioned to offer this service to podcasts due to our AI-powered content analysis technology, our engaged community, and our role within DSPs [demand side platforms]Marco Paglia, the company’s chief product officer, told TechCrunch via email:
One of the company’s goals, he added, is to become an authorized transcription provider for other services – just like its offerings in the lyrics space.