What We Saw at Data Natives 2017

Marta Lamela Casanova

d&a blog

Last November I attended the Data Natives Conference, in Berlin. They define themselves as “A conference for the data-driven generation” and it is absolutely true. I met a very diverse group of people: industry specialists, entrepreneurs, state-of-the-art startups, techies, data scientists and university professors.

Workshops day

 

It took place in an amazing location, the WeWork Sony Center. I chose two different workshops that seemed very interesting (and they were!):

 

  • Artificial Intelligence for Enhancing Customer Experience: An agile/design thinking workshop where Agnieszka Walorska, co-founder of Kickstart.ai, made us translate the customer’s needs into possible AI-applications, in a practical way.

 

  • Some Atypical Applications of Typical Machine Learning Algorithms: By Kirk Borne (@KirkDBorne), an astrophysicist who spent nearly 20 years supporting NASA projects. Kirk presented a variety of atypical use cases and applications (in science and business) of some typical textbook machine learning algorithms.

 

If you are interested in his presentation, you can download his slides here:

2-day conference

 

There were more than 50 speakers divided into two tracks. I spent most of the time in the Tech Room, with speakers talking about technologies Big Data, Machine Learning, AI and IoT.

Track 1: The Business of Tech

I attended a thoroughly exciting “Battle of Startups” where 8 startups presented their amazing projects. You should have them on your radar, I am positive they will be famous sooner rather than later!

Onwrks: A Smart Energy Solutions company. Their presentation was about how they have solved the issue of yaw misalignments on turbines using Time Series and Deep Learning.

Searoutes: They have created the equivalent of a Google Maps for ships using Machine Learning and Graph Theory, based on real shipping routes from around the globe and integrating weather information. They have a free API.

BiQMap: Presented their easy-to-use tool to create maps, built for journalists and web publishers. Because maps are beautiful!

Zoe: A group of 16-year-olds have created Zoe, a chatbot to help other teenagers with their school subjects. You can ask her about definitions, formulas, examples… or simply chat with her, she will reply in a “teenager way”, facilitating communication. The new generations of techies are coming strong!

Track 2: Tech Trends

Alan Akbik from Zalando Research talked about their work in NLP Processing & Text Analytics of Multilingual Text Data.

Also, Philipp Kähler from myTaxi explained how their Taxi Fare Prediction Model works. He also detailed the data infrastructure at myTaxi and how data is processed and stored, which is the foundation of every machine learning application.

And Dr. Fabian Werner from Lufthansa Industry Solutions talked about their Flight Duration Model, whose purpose is the prevention of delay. His talk focused on the hyperparameter optimization that they are doing with Bayesian Model Optimization (Gaussian Processes).

There were also talks about positively interesting topics (and more bank-related) like Fraud Detection and Prevention with ML, by Ralf Klinkenberg from RapidMiner. He explained how to detect Fraud Patterns, Anomalies, and Unusual Behaviors.

All the presentations are available on the following link

See you in Data Natives 2018!