AI2FUTURE / 2 unConference days / 16 speakers / 4 workshops

October 05th – 06th, 2017, Zagreb, Croatia

Manos Tsakgias

I love designing search, recommendation, and predictive analytics engines. I strive for products that scale by intersecting the latest theoretical advances with elegant engineering.

I hold a Ph.D. in Computer Science, an M.Sc. with Merit in Informatics, and a B.Sc. in Physics. My work focuses on improving conversion and user engagement through modeling of online user behavior–i.e., machine learning and information retrieval. I am happy to have worked together with great people at Sponsorpay, and at Yahoo! Barcelona on exciting large scale problems.

I am currently running 904Labs, an Amsterdam-based company that offers the first and only commercially available self-learning search engine. At 904Labs we bring state-of-the-art machine learning technology, that is accessible by only a handful corporations in the world, to the masses. If you run an online business, I strongly recommend you to checkout 904Labs product and get in touch; it will help you increase conversion rates and long-term user engagement.

Before 904Labs, in 2005, I co-founded FifthLimb in the U.K., an ERP software for recycling logistics; more than a dozen companies are currently running on it. In 2007, I joined University of Amsterdam as Ph.D. candidate and later on as postdoctoral researcher to work on search engines and online user behavior, topics on which I continue working today at 904Labs.



I have worked on a wide spectrum of information access problems ranging from predictive analytics and content tracking to entity linking and online learning to rank. Most of  my research has been conducted during 2007–2014 at Information and Language Processing Systems (ILPS) at the University of Amsterdam and it continues today at 904Labs, but at a lesser publication rate.

Currently I work at 904Labs: We are an Amsterdam-based Artificial Intelligence company offering next-generation search engines to e-commerce sites. We are a no-nonsense company, turning scientifically proven methods for onsite search into efficient, fully-functional, and easy-to-integrate solutions.




Follow Jan




Sessions and Workshops

There are many sessions and workshops at the conference, please select the ones that are great for you, but please also reserve some time for networking and meeting other great people.


Self Learning Search Engines

Ranking is at the heart of all search applications, whether the search concerns documents, entities, recommendations, or conversations. Today, we are witnessing an important shift towards online weakly supervised approaches: using natural interaction data for developing new rankers, for optimizing combinations of large numbers of rankers, and for the evaluation of rankers. These approaches have proved very powerful as they are robust to noise and can learn online. In the talk Manos will illustrate the principle of self-learning search systems, describe the current challenges, and share insights on the next generation of self-learning search engines.

Read more