Above: Oscar Celma. Photo credit: Flickr user Thomas Bonte.

Demian Bellumio is COO of Senzari, a digital music company, and a partner at MIA Collective, which is hosting the MIA Music Summit, a slate of events March 22-24 that brings together leaders from the music industry with Miami’s emerging startup community. He recently asked a few questions of Oscar Celma, a judge of the MIA Music HackDay and one of the speakers for the summit, about what the gathering means for Miami and the music industry. Celma is the new director of research at Pandora and a former senior research engineer at Gracenote. He also served as chief innovation officer at Barcelona Music and Audio Technologies (BMAT). Knight Foundation supports the summit, which will feature events at New World Center in Miami Beach and The LAB Miami in Wynwood. 

First of all, congratulations on the exciting news that you just joined Pandora as director of research.  

O.C.: Thank you very much! I’m really excited about my new gig at Pandora.

How do you think this new role will compare with your experience working at BMAT and Gracenote?

O.C.: Well, I’d say the main difference is that both BMAT and Gracenote are B2B companies, whilst Pandora is B2C. I learned a lot at BMAT, as we were just a few people when I started back in 2008. It’s awesome to be part of a small startup and see how it grows and becomes a mature company after all these years. I’m so happy to see that they are doing so well right now, given the current [economic] situation in Spain.

At Gracenote, a much larger organization, I learned how to effectively interact and work with people from different departments. I left so many friends at Gracenote, but I’m sure we’ll be in touch for a beer or two every now and then because I’m still an [Oaklander]

Now, my new position at Pandora is like a dream come true. This role combines both my background as a scientist and engineer/manager. Also being able to work with a group of very talented and smart scientists is so gratifying.

Last but not least, B2C is just another world; you really focus on your users, and try to improve the product day by day to provide them a better experience.

As a speaker at MIA Music Summit’s “Big Data and Music” panel, can you give us a preview on what you hope to discuss on it?

O.C.: I think that we’re still in the early stages of really exploiting and making a smart use of music big data. Till now we have mainly focused on the technologies and tools to process all this information. Still, we haven’t got that far in mining this vast amount of data and extract relevant and meaningful information from it. In principle, more and more data is always welcome, but we need to apply the science and models to extract some conclusions, which can be difficult sometimes.

I hope that we can also devote some time to discuss about the future of big data and its impact in the music domain. All in all, I’m sure it will be an interesting panel for most of the MIA attendees!

You wrote your book “Music Recommendation and Discovery: The Long Tail, Long Fail, and Long Play in the Digital Music Space” more than four years ago; what has changed since and what chapter would you add now?

O.C.: Related with the previous question, I think that during this period of time applied research in the music domain has benefited from the big data momentum. The main algorithms and ideas are more or less the same, but there are many more signals to process, and probably more sources than ever to combine. Nowadays people feel like music recommendation is a commodity and are more used to it, whilst a few years ago [it] was more difficult to sell this idea to them—or even to other companies that would clearly benefit from it.

I would add a chapter that’s more practical. A chapter that gives some tips based on what I empirically learned since the book was written and I moved to the industry. I think this could be useful for both the graduate student and the practitioner that’s already working in the industry. Writing about all this accumulated experience, on what worked and what didn’t worked—and the reasons why—could be very useful to prevent other people making the same mistakes I did. 

As you know, the MIA Music Summit aims to connect the startup and music worlds in Miami to foster digital music entrepreneurship. What city should we look at as a model for success in this space and why?

O.C.: Wow. That’s a very difficult question! I guess a cliché answer would be Silicon Valley. But depending on the type of entrepreneurship model you’re looking for, you can also turn to Europe where there are important hubs, like Berlin, that have played an important role in the music tech space, amongst others.

You have been at many music hackathons and will be selecting the five MIA Music HackDay finalists on Sunday, what advice do you have for the hackers looking to move on to the finals?

O.C.: Well, I think that the main point is to have fun whilst hacking. Think on a project that you’d enjoy working on for the whole weekend in sunny and beautiful Miami! (Not easy, uh?) That BIG IDEA that you think can change the music space, or maybe just the one that scratches your own itch. Once you get overexcited try to focus on just one killer feature, and implement it. Remember that you’ll only have 24 [hours] to create something. Finally, prepare and practice your pitch beforehand, and be ready to demo your hack to the audience in the most interesting way you can think of.

What do you hope to accomplish or learn at the MIA Music Summit?

O.C.: I hope to meet and get in touch with the music tech scene in Miami. Last time I was there in 2012 I met the cool guys from Senzari, and it seems that since then a lot of new and interesting startups have been created there.