Company founder Michael Welge, leader of the Automated Learning Group (ALG), National Center for Supercomputing Applications (NCSA), met Prof. J. Stephen Downie at the University of Illinois in 2002. They came together to collaborate on getting grants to use a data-mining test environment, called Data-to-Knowledge (D2K), to build a framework for music audio analysis and evaluation. The partnership proved successful as Downie and Welge were funded by both the National Science Foundation and the Andrew W. Mellon Foundation to make their vision for harnessing the power of advanced signal processing and machine learning a reality. They named their innovative analytic system, M2K (Music-to-Knowledge) . Over time, Prof. Downie's lab developed M2K into a state-of-the-art, open-source music analysis toolkit, in partnership with Kris West from the University of East Anglia and Sun Microsystem Laboratories.

Meanwhile, future partner, David Tcheng, a member of Welge's ALG team, along with Andreas Ehmann (Electrical Engineering), developed a set of real-time music genre detectors using the M2K toolkit. Andreas, a digital signal processing expert, developed the set of audio feature extractors, and David, a machine learning expert, developed a sophisticated variant of a fast decision tree learning algorithm to create their remarkeably accurate genre prediction models. Their system was affectionately named "Blinkie Thing" because of how the multicolored bars rose and fell to the music as it predicted the genres of the incoming music. "Blinkie Thing" was demoed several times at various academic conferences before Rob Schultz, Managing Partner of Illinois Ventures (www.illinoisventures.com) became aware of this exciting new technology. Schultz was inspired by the success and capabilities of M2K, began exploring the possibility of Illinois Ventures investing seed funding in a start-up. Illinois Ventures had previously invested in another data-mining company, River Glass, and recognized that Music and Media would be an excellent vertical market for specialized data-mining techniques.

Strongly encouraged by Welge (who knew Schultz), David Tcheng used months of accumulated vacation days to take a sabbatical to develop One Llama's first set of prototype music classification and recommendation systems. These prototypes were impressive enough to assemble a team and secure the first round of seed funding from Illinois Ventures. A founders team of top flight computer scientists and engineers were recruited from among the University community and they went to work, building both acoustic analysis and unique collaborative filtering platforms based the latest state-of-the-art advancements in audio analysis and machine learning. The result was one of the first, truly effective music similarity search and recommendation systems which was launched to the public about 11 months later in 2006.

The name One Llama comes from Rob Schultz and David Tcheng whose office windows overlooked the experimental farms at the University of Illinois. At the experimental farms was a flock of sheep (with orange and blue markings) that was guarded by a solitary llama. The llama was sociable and liked to pose with people for pictures!