Microsoft Azure (Microsoft) Case Studies AccuWeather's Transformation: Predicting Business Impact of Weather with Azure Machine Learning

Edit This Case Study Record

	AccuWeather's Transformation: Predicting Business Impact of Weather with Azure Machine Learning Microsoft Azure (Microsoft)

AccuWeather's Transformation: Predicting Business Impact of Weather with Azure Machine Learning

Microsoft Azure (Microsoft)

Technology Category	Analytics & Modeling - Machine Learning Application Infrastructure & Middleware - Database Management & Storage
Applicable Industries	Education Equipment & Machinery
Applicable Functions	Logistics & Transportation Sales & Marketing
Use Cases	Inventory Management Predictive Maintenance
Services	Cloud Planning, Design & Implementation Services Data Science Services
Challenge	AccuWeather, a leading global provider of weather forecasts, was faced with the challenge of providing custom weather-impact predictions for business customers. Businesses wanted more than just weather data; they wanted predictions of how the weather would affect their operations. Initially, AccuWeather provided this service on an ad hoc consultative basis, which was labor-intensive. Customers would share their sales history data, and AccuWeather would have a data scientist clean the data and create a custom model. However, this process was not scalable. AccuWeather wanted to automate this service, allowing customers to upload their data online and generate an automated prediction. However, they were concerned that the analytical tools they had at the time wouldn’t allow them to provide high-quality predictions. Read More
About Customer	AccuWeather, Inc., based in State College, Pennsylvania, is a leading global provider of weather forecasts. It provides minute-by-minute forecasts for 2.3 million locations around the world for more than 100 parameters—including temperature, humidity, rain, snow, and ice—for every hour over the next 90 days and every minute for the next two hours. AccuWeather gathers hundreds of thousands of real-time weather observations from land, ships, aircraft, satellites, and radar and crunches that data using its patented AccuWeather Forecast Engine. More than 180,000 websites, 200 television stations, 900 radio stations, and 600 newspapers feature AccuWeather forecasts. In total, nearly two billion people worldwide rely on AccuWeather to help them plan their lives, protect their businesses, and stay safe. Read More
Solution	AccuWeather turned to Microsoft Azure to build an automated, scalable weather prediction service for business customers. They chose Azure for its solid infrastructure as a service offering, and its rich and rapidly advancing platform as a service portfolio, including big data, machine learning, and AI capabilities. AccuWeather first moved its API business to Azure, which provided a more scalable, cost-effective way to manage the service. They then moved their big data storage and processing, using services such as Azure Blob storage, Azure Data Factory, and Azure SQL Database. To create highly accurate forecasts, they used Azure's machine learning tools, which were powerful yet customizable using R and Python code. In May 2017, AccuWeather launched AccuWeather D3 Express, a cloud-based analytics product that quantifies the impact of disruptive weather on a business. They then developed D3 Advanced, which delivers even more customized predictions using machine learning. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Machine Learning

Application Infrastructure & Middleware - Database Management & Storage

Applicable Industries

Education

Equipment & Machinery

Applicable Functions

Logistics & Transportation

Sales & Marketing

Use Cases

Inventory Management

Predictive Maintenance

Services

Cloud Planning, Design & Implementation Services

Data Science Services

Challenge

AccuWeather, a leading global provider of weather forecasts, was faced with the challenge of providing custom weather-impact predictions for business customers. Businesses wanted more than just weather data; they wanted predictions of how the weather would affect their operations. Initially, AccuWeather provided this service on an ad hoc consultative basis, which was labor-intensive. Customers would share their sales history data, and AccuWeather would have a data scientist clean the data and create a custom model. However, this process was not scalable. AccuWeather wanted to automate this service, allowing customers to upload their data online and generate an automated prediction. However, they were concerned that the analytical tools they had at the time wouldn’t allow them to provide high-quality predictions.

About Customer

AccuWeather, Inc., based in State College, Pennsylvania, is a leading global provider of weather forecasts. It provides minute-by-minute forecasts for 2.3 million locations around the world for more than 100 parameters—including temperature, humidity, rain, snow, and ice—for every hour over the next 90 days and every minute for the next two hours. AccuWeather gathers hundreds of thousands of real-time weather observations from land, ships, aircraft, satellites, and radar and crunches that data using its patented AccuWeather Forecast Engine. More than 180,000 websites, 200 television stations, 900 radio stations, and 600 newspapers feature AccuWeather forecasts. In total, nearly two billion people worldwide rely on AccuWeather to help them plan their lives, protect their businesses, and stay safe.

Solution

AccuWeather turned to Microsoft Azure to build an automated, scalable weather prediction service for business customers. They chose Azure for its solid infrastructure as a service offering, and its rich and rapidly advancing platform as a service portfolio, including big data, machine learning, and AI capabilities. AccuWeather first moved its API business to Azure, which provided a more scalable, cost-effective way to manage the service. They then moved their big data storage and processing, using services such as Azure Blob storage, Azure Data Factory, and Azure SQL Database. To create highly accurate forecasts, they used Azure's machine learning tools, which were powerful yet customizable using R and Python code. In May 2017, AccuWeather launched AccuWeather D3 Express, a cloud-based analytics product that quantifies the impact of disruptive weather on a business. They then developed D3 Advanced, which delivers even more customized predictions using machine learning.

Impact #1	By developing D3 Advanced in Azure, AccuWeather was able to get to market faster than it could have if it had developed its own machine learning services. The company’s data scientists were able to focus on creating great foundational technology that delivers a scalable solution for many customers, freeing up time to develop new products and provide even more tailored consulting services. AccuWeather sees data and analytics as a huge growth area and has made it a major initiative to increase its ability to provide more insights from D3. By using Azure Machine Learning, they can meet a variety of customer needs, learn from all the analytics, and put that learning into new products and services to improve their offerings. Moving forward, AccuWeather will conduct all its data and analytics product development and hosting in Azure, and is moving all its products out of AccuWeather datacenters and into the cloud.

Impact #1

By developing D3 Advanced in Azure, AccuWeather was able to get to market faster than it could have if it had developed its own machine learning services. The company’s data scientists were able to focus on creating great foundational technology that delivers a scalable solution for many customers, freeing up time to develop new products and provide even more tailored consulting services. AccuWeather sees data and analytics as a huge growth area and has made it a major initiative to increase its ability to provide more insights from D3. By using Azure Machine Learning, they can meet a variety of customer needs, learn from all the analytics, and put that learning into new products and services to improve their offerings. Moving forward, AccuWeather will conduct all its data and analytics product development and hosting in Azure, and is moving all its products out of AccuWeather datacenters and into the cloud.

Benefit #1	AccuWeather receives more than 40 billion API calls a day for weather information.
Benefit #2	AccuWeather provides forecasts for 2.3 million locations around the world.
Benefit #3	Nearly two billion people worldwide rely on AccuWeather forecasts.

Benefit #1

AccuWeather receives more than 40 billion API calls a day for weather information.

Benefit #2

AccuWeather provides forecasts for 2.3 million locations around the world.

Benefit #3

Nearly two billion people worldwide rely on AccuWeather forecasts.

Download PDF Version

Overview

AccuWeather's Transformation: Predicting Business Impact of Weather with Azure Machine Learning

Operational Impact

Quantitative Benefit