In the world of AI, ML and Big Data analysis, we have forgotten about our main clients - people who are not interested in querying databases or waiting for the result of data preparation - they want to easily play with data themselves.
In this presentation, I will explain to you:
- what are analytical cubes
- who will use them and how can they do it
- what are the differences between Apache Kylin and Microsoft Analysis Services and how to prepare a cube in these environments

#kylin #analysisservices #analyticalcubes #ssas #apache

1) When Data Lake becomes a Data Swamp
2) Data access patterns and sharing - what could possibly go wrong?
3) Why do we need to introduce Data Layers
4) Bronze / Silver / Gold Data Layers and the concept behind it
5) How to make it work - tips and tricks.

#data_swamp #good_practices #sharing_data #thinking_about_users #data_lineage

- Where is entity resolution and deduplication needed
- Why is it a tough problem to solve
- Demonstration of open source technologies to solve this

#entityresolution #recordlinkae #deduplication #dataquality #ml

• Microservices
• Industrialized data ingestion
• Data virtualization
• Data privacy
• Operational data store

#Kubernetes #microservices #spark #datalakehouse

-How to transform from building one deep learning model per month into evaluating and deploying hundreds of them in a single week?
-Building MLOps solution with CICD practices using CDK
-How to detect underground structures based on a bunch of radar signals and no labels?
-Should we avoid manual steps in the automatic Machine Learning pipeline?
-Can we use lambda aliases to differentiate between dev and prod environment?

#MLOps #CDK #DeepLearning #Automatisation #Python

Watch this short 10 minute video, see how technologies like Kafka, .NET core, Scala, Java are used in AdTech and learn more about the journey of data in AdTech based on Adform example. The presentation goes through 5 V’s of Big Data (volume, variety, velocity, veracity, value) and shows where does all this data come from as well as gives a glance at the data pipelines.

#AdTech #BigData #Adform #DataPipelines #kafka #scala #.NET #java

Traditional data platforms are complex, expensive and time consuming to provide a business value. Why not to consider a solution which your business users would love to use? Why a typical analytical project have to take months to deliver?
Let me show you, how a single, fully self-service platform can provide all services orchestrated together to address your analytical needs in action!

#analytics, #bigdata, #ROI, #datadriven, #selfservice

AI is as much a management challenge as it is a technological challenge.
There has been, and still is, a big debate about data driven management. We've long been encouraging managers and decision makers to use whatever data they have available, to draw insights and to make more informative and verified decisions. As a result we can observe an increase in Data Analytics and Data Science employment. What we can also observe is increasing complexity of systems, tools, applications, regulations and policies around their day-to-day work.Now is the time to encourage everyone to think about how to organize and manage all of the teams, tools, pipelines, and applications that we've brought to life in Data Science.

#ai, #strategy, #mlops, #data science, #ml, #management

During the presentation we will show how we delivered project of implementation of Augmented Analytics platform for one of our clients. We will describe what were the client’s challenges and how the platform was able to solve them.

#Business Intelligence, #AI, #Artificial Intelligence, #Augmented Analytics

Data preparation is a key phase in any machine learning/data mining process. In most cases is complicated, undocumented and there are serious issues in moving all process from the development environment to production environment. I want to show that, with the right tools, it does not have to be a problem.
Ab Initio is a platform that allows for processing, transforming and managing data in every single aspect, from low level details to high abstraction concepts. The core of that process are graphs, which are responsible for data transformations. They provide declarative-streaming approach to data processing. During short presentation I will show how to load the data, process them into required form, and use it to build ML model. This approach is so flexible that will be useful for many people with different technical skills. Programmers can use graphs as a normal programming language, data scientists can operate on data without touching programming details and managers can understand the whole process and have full documentation of it.

#AI/ML #mlops #datamining #ETL #declarativeprogramming