Company based in Italy Musixmatch known for providing community-supported lyrics to major music streaming platforms, including Spotify, Apple Music, YouTube Music, Amazon Music and Tidal. It’s coming out now a new platform for podcasts combines AI-generated transcription and community-verified editing.
While there are millions of podcast shows and episodes available to listeners, Musixmatch thinks podcast search is broken. Thus, it shows a lot of great podcasts that are not connected to potential fans. So they are using their experience in training AI models through lyrics and leveraging their NLP (natural language processing) expertise to improve transcription, search, discovery. and share podcasts.
Musixmatch’s podcast platform automatically generates daily transcriptions of some of the top podcast episodes on different topics and charts. It uses the NLP base model architecture, Umberto, to tag keywords like places, people, and topics with Wikipedia IDs – alphanumeric IDs associated with topics on Wikipedia. (For example, this link indicates the Wikipedia ID associated with TechCrunch.)
Because of this approach, it says people searching for these topics in any language will get the correct results.
The startup explains to TechCrunch that based on these IDs, it creates a topic graph called TopicRank, which ranks podcasts based on factors like mentions in an episode or people’s expertise topic presentations – improve search results for podcasts when users are searching for related topics.
“Thanks to this categorization, people can finally search for any particular keyword and find transcribed podcasts that match their query, sorted by their relevance. Our search index returns a much more detailed and in-depth range of results than any other listening service based on standard RSS metadata and predefined categories and categories, the company stated.
When a user searches on Musixmatch’s podcast platform, it shows excerpts from transcriptions that mention the searched term. If they click on the result, the podcast will start streaming from the timestamp of the snippet referring to the phrase. That’s pretty neat when you need to listen to a few minutes of audio while studying something.
Musixmatch has long relied on its community to make precise edits to the lyrics, and now they ask these users to do the same with podcasts. The company’s new podcast portal also includes a tool called Podcast Studio, which allows podcast editors and owners to edit AI-generated transcriptions – especially useful for things like people and brand names trademarks or cultural references.
If there is no recording for a particular episode, owners or community members can use Podcast Studio to create the recording. Musixmatch says it takes about five minutes for the AI to create a recording for an episode. Regular listeners can also donate an episode to the transcript so that the community prioritizes those episodes.
It’s important to note that on Musixmatch’s platform the AI-generated transcriptions will have tags like “Speaker 1” and “Presenter 2”, while Community-edited episodes will have labels with the names of speaker – along with a “verified” label.
The company is also making sharing easier by displaying tags with text from podcasts with a shareable link. Furthermore, it is working on a feature called audiograms, which are small shareable videos that include audio and scrolling texts from a podcast.
Musixmatch doesn’t want to keep all this data to itself. It allows podcast owners to publish audio recordings to their web and app feeds. And since these texts are SEO-friendly, it argues that it will make it easier for listeners to find them.
Some of Musixmatch’s partners that it says are using its tools for transcription include The Financial Times’ “The Talent Show”; “Beyond the Ordinary” and “Why I Run” by Red Bull; and the entire Chroa Media production process.
While Musixmatch’s podcast platform offers features for listeners, it’s not trying to be a podcast player. The startup argues that its competitors are companies that work in audio analytics – including apps that provide transcription services (such as Old castle).
“We think audio analytics (AI, semantics, etc.) will be a must-have in the near future, for a variety of use cases. We’re in a unique position to provide that service for podcasts because of our AI-powered content analytics technology, our engaged community, and our role in the DSP. [demand side platforms]that we’ve made available to third-party content,” the company’s chief product officer, Marco Paglia, told TechCrunch via email.
He added that one of the company’s goals is to be a verified sound recording provider for other services – like theirs in the lyrics space.