Developing and Deploying a Churn Prediction Model with Azure Machine Learning Services

Patty Ryan January 10, 2019 Jan 10, 2019 01/10/19

Business Problem For a subscription service business, there are two ways to drive growth: grow the number of new customers, or increase the lifetime value from the customers that you already have by retaining more of them. Improving customer retention requires the ability to predict which subscribers are likely to cancel (referred to as churn...

Real-Time Time Series Analysis at Scale for Trending Topics Detection

Omri Mendels January 2, 2019 Jan 2, 2019 01/2/19

Managing big cities and providing citizens public services requires municipalities to have a keen understanding of what citizens care the most about. And, as the cities we live in seek more and more to become the 'smart cities' of tomorrow, this means gathering and analyzing vast amounts of data. Data gathered from social and municipal sources...

Improving Safety and Efficiency in BMW Manufacturing Plants with an Open Source Platform for Managing Inventory Delivery

Ryan Pedersen December 19, 2018 Dec 19, 2018 12/19/18

Background German car manufacturer BMW has always been at the forefront of technological advancement within the auto industry. And a major part of the company’s innovation focus is on its production system, ensuring quality and flexibility alike. When it comes to cutting-edge production systems, and more precisely logistics, new ...

Social Stream Pipeline on Databricks with auto-scaling and CI/CD using Travis

Mor Shemesh December 12, 2018 Dec 12, 2018 12/12/18

Background For the tech companies designing tomorrow’s smart cities, making local authorities able to collect and analyze large quantities of data from many different sources and mediums is critical. Data can come from different sources – from posts on social media and data automatically collected from IoT devices, to information ...

Active Learning for Object Detection in Partnership with Conservation Metrics

Olga Liakhovich November 6, 2018 Nov 6, 2018 11/6/18

Introduction Using artificial intelligence to monitor the progress of conservation projects is becoming increasingly popular. Potential applications range from preventing poaching of endangered species to monitoring animal populations in remote, hard-to-reach locations. While proven to be extremely effective, computer vision AI projects ...

How to Build A K8S Http API For Helm, and Serve Micro-services Using A Single IP

Itye Richter October 25, 2018 Oct 25, 2018 10/25/18

Image credit: Pexels.com Background Axonize is a global provider of an IoT orchestration platform which automates the process of IoT deployments, cutting the process down from months to days. Their innovative platform consists of Data Gateways, which can accept even non-standard IoT sensor input, reformat the payload, and then deliver ...

Attaching and Detaching an Edge Node From a HDInsight Spark Cluster when running Dataiku Data Science Studio (DSS)

Amy Boyd August 20, 2018 Aug 20, 2018 08/20/18

Background Microsoft Global Partner Dataiku is the enterprise behind the Data Science Studio (DSS), a collaborative data science platform that enables companies to build and deliver their analytical solutions more efficiently.  Microsoft has been working closely with Dataiku for many years to bring their solutions and integrations ...

Infrastructure as Code – On demand GPU clusters with Terraform & Jenkins

Tomer Rosenthal August 9, 2018 Aug 9, 2018 08/9/18

Background Developing robust algorithms for self-driving cars requires sourcing event data from over 10 billion hours of recorded driving time. But even with 10 billion-plus hours, it can be challenging to capture rare yet critical edge cases like weather events or collisions at scale. Luckily, these rare events can often be simulated ...

Unsupervised driver safety estimation at scale, a collaboration with Pointer Telocation

Omri Mendels July 30, 2018 Jul 30, 2018 07/30/18

Background Connected cars, which have the ability to collect and transfer data via cellular networks, offer us the opportunity to make road transport safer, more organized and more ecologically friendly. To push forward widespread adoption and help the technology fully enter the mainstream, manufacturers and third parties, like insurance ...

Semantic Segmentation of Small Data using Keras on an Azure Deep Learning Virtual Machine

Margaret Meehan July 18, 2018 Jul 18, 2018 07/18/18

Introduction As previously featured on the Developer Blog, golf performance tracking startup Arccos joined forces with Commercial Software Engineering (CSE) developers in March in hopes of unveiling new improvements to their "virtual caddie" this summer. Powered by Microsoft Azure, Arccos' virtual caddie app uses artificial ...