Interview: A Look into Mendeley's API Development

The Mendeley API provides access to research tools and data. In the interview below we discuss with Rosario García de Zúñiga, Senior Software Engineer & Team Leader at Mendeley, how she sees the future of open APIs. This interview is part of a series conducted by ProgrammableWeb and Hojoki (where I work), an aggregation tool for apps such as Google Drive, Dropbox and Evernote, that helps support team work.

Who are you and what do you do?
My name is Rosario García de Zúñiga and I am a software engineer at Mendeley. I have been working in our API since its beginnings, about 3 years ago.

How important is the API for your product? And why?
The API is a key part of our product. It opens all the data we have available for third party developers to create awesome applications and make research more accessible. It also allows developers to create interesting applications that we wouldn't have time for.

Seeing Twitter restricting their API, do you have similar plans at Mendeley?
We don't have plans to make our API more restrictive. The more users who have access to our research data the better. We believe in open access and collaboration so we’re actually planning to add more API methods and expand the possible use cases of our data.

What are your predictions for the future of open APIs in general? Will they decline due to others following Twitter’s path or will Twitter’s approach be proved wrong?
We've seen a huge growth in open APIs and this results in some really great applications and we expect it to continue.

Who do you see as innovators when it comes to pushing the possibilities with what can be achieved with APIs?
OpenSNP is a great example: who would have thought a few years back that you could just upload your genome and find the latest research and help researchers discover new genetic associations? That is just awesome.

Mendeley’s API provides not only access to your app’s functionality but to a huge data set of scientific data. Do you plan to monetize this data?
Yes, we do. We’re already seeing paid apps emerging in the Mendeley ecosystem. In the future, we want to make it easier for developers to monetize the apps they’ve built with our data, so there will be monetization opportunities for us, too.

You recently hosted a competition for 3rd party developers to develop scientific apps based on Mendeley. How did this go?
Our main goal was to draw attention to the possibilities of developing research apps using our API and kickstart the Mendeley ecosystem. As such, it was an amazing success: We had good press coverage and the judges - people like Tim O’Reilly and Werner Vogels, the CTO of Amazon - were extremely impressed with the quality of the apps delivered. During the contest, more than 40 new apps were developed, and the number of API calls went from a few thousand per month to more than 20 million per month! Right now, there are more than 260 third-party apps and more than 100 million API calls per month.

Technically speaking, what’s your API related roadmap for the near future? What’s to come?
The roadmap is looking very exciting. Our main goal is to provide an API that allows developers to create fully functional clients like the one we provide. In the short term, we are looking at using OAuth 2, which has recently been finalized and is easier to use than OAuth 1. We are also looking to improve services that are popular with API users and important to the clients we build, particularly in the area of document discovery.

Be sure to read the next Best Practices article: Interview: Viralheat CEO Raj Kadam on API Growth