craftworks Tech Talks

CI/CD for Machine Learning

In software development, Continuous integration and delivery (CI/CD) is widespread and used to iterate faster. Machine Learning Models depend however not only on the code but also on data and other parameters. Thus, the topic of CI/CD is more complex for ML models. In this talk, you will get an introduction into CI/CD for machine learning, open-source tools, and one on-premise use case with Jenkins. 

What you will learn:

• What are the challenges of CI/CD for ML especially
• What tools can be used to handle data and retraining with CI/CD
• Concurrent usage of on-premise hardware
• One practical example of a real-world application

Target Audience: Data Engineers, Data Ops, Data Scientists

Speaker:

Intro to Spark and Databricks

This Tech Talk will give you an introduction to Spark and how to write and execute Spark code with Databricks and what it is needed for. Apache Spark is an open-source distributed general-purpose cluster-computing framework. Currently its the leading platform for large-scale SQL, batch processing and stream processing. Azure Databricks is an Apache Spark-based analytics platform and accelerates innovation by unifying data science, engineering and business. Join this session if you are interested to get to know Spark and Databricks better and how to use it.

What you will learn:

• What is Spark and when do I need to use it?
• Databricks Basics and how to setup Azure Databricks
• How to write basic PySpark code
• PySpark & Databricks in action: demonstrated with a use case

Target Audience: Data Engineers, Data Ops, Data Scientists

Speaker:

UI/UX for industrial settings

While great attention is paid to the user interface in the B2C area, this is often neglected in industrial applications. However, usability is a success factor for industrial solutions as well: The work with well-designed interfaces is more profitable due to faster error prevention and fewer operating errors. 

In this session, we discuss a human-centered design process for building industrial software solutions that in the end can be used smoothly, effectively, and efficiently. Further, we will demonstrate what usability can look like in real-world use cases. 

What you will learn:

• Basic concepts and values of designing in a user-centered approach. 
• How a user-centered design approach can be followed when building industrial solutions. 
• Examples of real world use cases.

Target Audience: Product Owner, Designer, Software Engineer

Speaker:

SparkML

In this session, we will focus on Spark and how to use the framework for Machine Learning. We will walk you through the basic concepts and show you how Spark can efficiently perform data exploration, cleaning, aggregations, and train ML models.

In the tech talk, we will also guide you through a use case backing up the theory. 

What you will learn:

• Basic concepts
  • Transformers, Estimators, Pipelines
• Data cleaning and Feature engineering
• Pipeline training, selection and evaluation

Target Audience: Data Scientists, Data Engineers

Speaker: