Skip to Content
Author's profile photo Tammy Powlas

What’s new in SAP Predictive Analytics 3.3 Webcast Recap

This was an SAP webcast from last week.  More upcoming related webcasts can be found here:

Source: SAP

Source: SAP

This is an important release for SAP; build highly scalable models, work with other applications in SAP, deploy into customer’s environments and data privacy support

Above are the four main features

API for Python – mass script models

PAI – Predictive Analytics Integrator – solve end to end use cases for integration

Deploy Predictive on Linux – allow for standardization

GDPR support to comply with EU law

Automated Analytics

Source: SAP

Python scripting API

Source: SAP

Create more powerful workflows with Python

Building a loop through scripts

Python is open-source, well-adopted in data science community

Source: SAP

Source: SAP

Samples for classifications and clustering are provided

Source: SAP

You have to install Python; SAP does not do that

Source: SAP


What is new in Predictive Analytics Integrator

Source: SAP

How manage lifecycle of model once it is embedded in SAP application – that is what PAi does

Customers can build “their own”

Source: SAP

Predict end date of a contract

New tab for PAi in Predictive Factory

Source: SAP

Can install Predictive Factory on Linux with 3.3 release

Source: SAP

Data privacy; deletion of personal data; define length of time

Question and Answer

Q: When available?

A: 3.3 avail: the plan in November 13.

Q: Is there integration with Data Hub (current or planned)

A: Integration with Data Hub is planned

Q: Does PAI work with the SCP Workflow component?

A: for SCP workflow, you could use Predictive Services (REST APIS) for various scenarios like classification, recommendation etc..


Assigned Tags

      You must be Logged on to comment or reply to a post.
      Author's profile photo Antoine CHABERT
      Antoine CHABERT

      Great recap Tammy! Thanks for this.

      Author's profile photo Chris Gruber
      Chris Gruber

      For those of you who wanted to get more information on the Python samples, check out the following link:


      Author's profile photo nan keew
      nan keew

      Thank you. It’s fun to read  And we recommend some good articles