AccuWeather Showcases patented AI-based model that turns risky weather obstacles into business opportunities

AccuWeather’s Data Sciences and AI team will be showcasing their latest innovation in severe weather forecasting at this year’s Databricks Data + AI Summit. The team is also issuing a call to action for potential partners to leverage their third-party data to model specific, weather-related risks to their business.

“This innovation is a perfect example of what we do best and emphasizes our core mission,” said Chris Patti, Chief Data and Technology Officer at AccuWeather. “We combine unparalleled data and analytics with over 6 decades of human meteorological expertise. This allows us to help tell stories, put the weather in context, and provide clear and understandable assessments of what weather-related scenarios may unfold in given situations to help people and organizations manage risk and plan ahead.”

AccuWeather uses this patented and unique weather risk measurement approach to bring its rich and quality-assured weather data together with partner-provided data on the Databricks Data Intelligence Platform. The process uses machine learning to help users understand where and when the most significant impacts on their business will likely occur.

“Context is key in assessing weather impact. In Atlanta, two inches of snow paralyzed the city in 2014, whereas in Denver or Boston, that amount would be a mere nuisance. Understanding weather relevance at your location is critical for assessing weather-risk accurately,” said Tim Loftus, AccuWeather’s Manager, Data Science. “Integrating our expertise, enriched data, and our partners’ business insights with Databricks and their open-source tools like MLflow can drive transformative collaboration — from optimizing business processes, to saving lives and protecting property.”

For example, a home improvement retailer selling grills may lose money when bad weather strikes on peak weekend business days. This AI-based process captures the weather-risk by combining that retailer’s business impact data with AccuWeather data in a proprietary machine learning algorithm to generate an impact value that can signal where and when overstocking will occur. With these custom weather insights, the retailer can plan to optimize store layout and product placement, or implement promotions and discounts on grills to draw more customers to their store and minimize risk.

Databricks users can take advantage of the AccuWeather Data Suite available on the Databricks Data Intelligence Platform and now use this new impact value module to gather more understanding of how weather is impacting their bottom line. Accuweather’s rich datasets are also available on the Databricks Marketplace including historical, forecast, and demand forecasting weather data.