26-28 November, 2019, Vilnius

Conference is over! See you next year.

Confirmed Talks

Jon Bratseth

Verizon Media, Norway

Talk

Big Data Serving: The Last Frontier. Processing and Inference at Scale in Real Time

The big data world has mature technologies for offline analysis and learning from data, but have lacked options for making data-driven decisions in real time. This talk introduces vespa.ai – a mature open source platform for storing, selecting processing and making model inferences over large data sets at end user request time.

Session Keywords

Big Data
Machine Learning
Search
Recommendation Systems

Kris van der Mast

VaHa, Belgium

Talk

Chat Bot Building Story - The Good, the Bad and the Unexpected

Since September 2016 I’ve been building bots and I want to share with you what we encountered during development, sometimes hilarious, sometimes not so much. Join me in this talk and gain insight on how to think about chat bots and how you can get started based on experience.

Session Keywords

bots
nlp

Markus Winand

winand.at, Austria

Talk

More Than a Query Language: SQL in the 21st Century

“Great News–The Relational Model is Dead” was a prominent comment on the release of the new SQL standard in 1999. The message behind the provoking statement was that SQL has evolved beyond the relational model. As much as this move was discussed at that time, it took decades until database vendors caught up with this idiomatic change. Many developers haven’t heard of it until today.

Session Keywords

sql
databases

Dionizas Antipenkovas

Peekdata.io, Lithuania

Talk

Embedded Analytics through Data API or What Big Companies are Hiding from You

There are more than 20 years of experience of building amazing ETL and BI solutions and tools. These solutions proved being very useful and will serve us in the future.
However, such BI ecosystem was egoistically targeted for closed use cases with a focus on internal needs of companies using it, so only large companies allowed themselves offering great Analytical tools for their customers as self-service. 

Session Keywords

Schema less Data API
Embedded Analytics

Bradley Arsenault

Electric Brain Software Corporation, Canada

#1 Talk

Best Practices for Building AI Datasets

Datasets are the most basic building blocks in AI systems, and the most innovative solutions often require manually collecting and labelling data. Yet most teams put their emphasis on magical models instead of solid, high quality datasets. In this talk, I will discuss many of the best practices for building AI datasets from scratch.

Session Keywords

Data Collection
Datasets
Machine Learning

#2 Talk

The Joy of Designing Deep Neural Networks

Deep Learning is the most powerful class of learning algorithms that we have ever invented. In this talk, I will try to share the joy that I felt the first time I designed a deep neural network, and how simple intuitions on neural networks have lead me to great designs and amazing accuracy.

Session Keywords

Deep Learning

Elena Lazovik

TNO, The Netherlands

Talk

Runtime Modifications of Distributed Big Data AI Models and Its Parameters

There has been phenomenal growth of interest in distributed data analysis with large software and other commercial entities. They are trying to construct the solutions that facilitate flexible, scalable and heterogeneous IT-infrastructures dealing with Big Data analysis. Distributed data processing platforms such as Hadoop or Apache Spark became a de-facto standard in the world of Big Data processing, and in cloud economy. 

Session Keywords

Distributed Computations
Big Data Processing
Apache Spark
Distributed Data Science

Vygantas Butkus

Danske Bank, Lithuania

Talk

The Journey Through Real Estate Price Model Development

Together with his team, Vygantas has developed a model that predicts the selling price of homes in Denmark. The development was an exciting journey with many challenges and different attempts to overcome them. Sometimes the build process is more interesting than the final model itself. Thus, Vygantas is very happy to offer you a walkthrough of this experience with all the pains and joys encountered along the way.

Session Keywords

Modelling
Machine Learning
Real Estate

Emanuil Tolev

Elastic, UK

Talk

Better Data Exploration and Visualisation with Elasticsearch and Kibana

Data science has incredibly sophisticated analysis tools which are very effective, but rarely very convenient for their operators. In this live demo we will focus primarily on the prototyping and exploration phase of a data scientist’s workflow.
We’ll push a slice of a bigger dataset held in pandas into Elasticsearch. We’ll then use Kibana to build on the fly visualisations and dashboards, allowing us to explore and discover the data’s features in a more intuitive, faster way than writing commands in a jupyter notebook.

Session Keywords

Data Science
Workflow
Visualisation
Datasets

Dr. Andreas Bühlmeier

Dr. Bühlmeier Consulting, Germany

Talk

Breakthroughs and the Future of (Deep) Reinforcement Learning

Due to recent successes in winning computer games against humans, Reinforcement Learning (RL)
received a lot of attention in the media.
This presentation will elaborate the foundation of RL and demonstrate how it is implemented.
Also, it will demonstrate how we can track and understand a system’s learning progress.
Most recent trends of RL research and applications in logistics and pharmacy will be discussed.

Session Keywords

Machine Learning
Reinforcement Learning