Prophesy 5.2 Release

January 21, 2020

With our latest release, Scintel continues to empower your team with a full suite of Data Science capabilities. Generate predictive models in hours, not weeks or months, and give your team more time to focus on the results that will take your organization to the next level.

What’s New 

• I/O datasets can be expanded vertically by adding missing samples or observations
• I/O pipelines now report totals for their outputs (row counts and file sizes)
• Multiple business value variables and metrics can be computed and stamped for feeds
• New analytical models (LightGBM and ensemble models)
• Insights and Influencers of feeds with micromodels (by each micromodel)
• Activation of micromodel-based feeds through Prophesy Runtime (additional web services)
• Activation of raw data feeds through Prophesy Runtime by automatically triggering defined I/O pipelines (additional web services)
• Support for raw datasets with separators other than comma
• Continuous UX improvements (new logo, archival of I/O pipelines, tooltips for Tuning parameters among several more)

Performance Improvements in 5.2

• Models can now be built in parallel to speed up the turnaround time further
• Reanalysis (Rerun Analysis) of feeds is up to 50% faster; huge time-saving in tuning and analyzing large feeds (25 million+ observations)
• Compatibility with API Gateways like NGINX to offload SSL termination for enhanced scalability

Deployment Improvements in 5.2

• Pre-built images for Amazon AWS and Microsoft Azure for faster onboarding of clients that are cloud-ready
• Flexibility in deployment with or without containers like Docker for Prophesy Runtime for low-latency turnaround for high volume activations

About Scintel Analytics

Scintel Analytics is a full-service analytics consulting firm focused on helping companies make better decisions faster. Prophesy, our predictive analytics platform, is used by our clients to solve a variety of business problems, such as customer churn, margin compression, predicting sales performance, etc. To find out how Prophesy can be applied to your business, email us at prophesy@scintel.com.