VentureBeat

SeeqPod is a music search and recommendation website that indexes uploaded music from around the web and lets you instantly play the songs you find when you search.

There are other music search sites with some similar functions, like imeem and Mercora, but SeeqPod is different because it pulls music from across the web, legal or not. Imeem depends on uploads to its site, and both it and and Mercora try to let users listen only to legally obtained music.

SeeqPod is one of the many companies seeking to benefit from the massive interest in music sharing online. There’s money here, too, like the $280 million CBS paid for Last.fm at the end of May.

The company builds its index by crawling music-related sites, blogs, social networks –anywhere that playable music might be found. When you go to the site, you see a list of songs that the crawlers are finding in real time. Next to the search bar are two options: “Search,” and “Discover.” Search does just what you’d expect and returns results that match the search term. Discover, on the other hand, will analyze the artist or song in the query and recommend songs and musicians that SeeqPod’s engine thinks you’ll like.

SeeqPod bills itself as “playable search,” which means that you can take the results pulled from across the web, put them in a list, and play them on the spot. You can also take this list and embed it via widget into your website, social network page or blog.

Combining the elements of music search, recommendation, and playability sets SeeqPod apart from recommendation engines like Pandora and Last.fm. While you can search for artists on Last.fm, for example, you can only listen to snippets of the songs you find. Functionally, SeeqPod is most similar to imeem; on both of these sites, you can search for artists and songs and build lists of songs you want to hear on the fly. Imeem’s recommendations, however, come directly from other people in the form of pre-made lists, and SeeqPod’s are automatically generated by its engine.

This engine also does things differently. Pandora deploys humans to identify the underlying characteristics of a song and link songs accordingly. Last.fm creates associations based the collective tastes of its users — so, if you like an artist that other people also liked, then Last.fm may recommend another artist those people liked, etc. SeeqPod, however, crawls the net for playlists, social network preferences, and other sites discussing music and relates songs based on the frequency with which they are listed or discussed together. Then, when you input an artist or song and click “discover,” SeeqPod offers you a list based on the associations it has made.

The technology was built in the Department of Energy’s Lawrence Berkeley Labs, and SeeqPod is making some pretty bold claims about its power. Co-founder Kasian Franks says that its type of contextual analysis mimics the way the human brain creates associations, and will work in any vertical market (health, finance, etc) the company decides to target down the road. What’s more, says Kasian, the technology’s ad-matching potential enormous, and he expects to eventually build an ad network or team up with one and take Google head-on.

We found the results to be solid, but not stellar, and occasionally off. Searching for recommendations based on the laid back lounge music of Thievery Corporation, for example, we got the significantly less mellow Rob Zombie among our suggestions. Also, the system didn’t recognize “Beatles” as a term and only returned recommendations when we added a “the.”

That being said, SeeqPod offers a different spin on music search and discovery, and there are a lot of people looking for ways to discover new music. It has a ready market if it turns out to be something useful

The company has raised an undisclosed amount in the high single-digit millions without going to any VCs.

.