Azure Machine Learning Services – Overview

We rely heavily on Microsoft’s cloud platform Azure during for our analytics workloads at the Austrian Postal Service. Azure has grown rapidly over the past few years and is adding features at a very fast pace, so it is easy to lose track which services are (still) offered and what services one should use . … Read more

Quick implementation of Yolo V2 with Keras!

Feb 22, 2018 I do not hold ownership to any of the above pictures. These are merely used for educational purposes to describe the concepts. Real time multiple object localization remains a grand debate in the field of digital image processing since many years. With the invent of Deep Learning and convolutional neural networks, the … Read more

Ordinal Logistic Regression

An overview and implementation in R Feb 19, 2018 Fig 1: Performance of an individual — Poor, Fair, Excellent Can you guess what is the common link in the variables mentioned below: Job satisfaction level — Dissatisfied, Satisfied, Highly Satisfied Performance of an individual — Poor, Fair, Excellent Impact of a regulation on bank’s performance — Positive, Neutral, Negative The variables are not only … Read more

Conway’s Law

Many organizations have become adept at identifying what they need from software development projects, based on a keen understanding of their business goals. Even so, they’re often surprised to find out that the end results don’t achieve the transformative impact they were expecting. Their mistake? Overlooking the importance of Conway’s Law. In 1967, Melvin Conway … Read more


Hi, I am Christoph, the Lead Data Scientist in the BI Competence Center at the Austrian Postal Service. I am responsible for designing the data science architecture, building the data science team and for coding up predictive models. Prior to joining the Austrian Post, I worked as a financial consultant at KPMG. I have a … Read more

Writing Custom Keras Generators

The idea behind using a Keras generator is to get batches of input and corresponding output on the fly during training process, e.g. reading in 100 images, getting corresponding 100 label vectors and then feeding this set to the gpu for training step. The problem I faced was memory requirement for the standard Keras generator. … Read more

Tips for Using Data to Solve Company Issues that You Can Master Today

It’s not enough to do data analysis (Credit rawpixel: Unsplash) Stop management from ignoring your analysis Feb 3, 2018 To a data analyst, there is nothing more exciting than the data revealing insights about real organizational issues. However, it is completely deflating to present the insights to management and accomplish nothing. From talking to other data … Read more