Scaling AI with Data Intelligence Webcast Recap

[TammyPowlas]

This was a recent webcast from SAP.

Please be sure to join us for these upcoming webcasts:

August 27 BI: SAP Analytics Cloud – Preparing your data

September 3 BI: Discover SAP BusinessObjects BI4.3

September 4 BI: SAP BW/4HANA Overview and Roadmap

September 5 BI: SAP Analytics Cloud - Data Connectivity, including Road Map

September 10 BI: SAP Analytics Cloud for SAP BusinessObjects Enterprise customers - Roadmap Update

September 11 - BI: SAP BW/4HANA Conversion

_________________________________________________________________________

What is Data Intelligence? Where Enterprise AI meets Intelligent Information Management



Source: SAP

What does this mean?

3 major capabilities - scaling AI, extract value from distributed data, embrace open technologies



Source: SAP

Jewelry company using AI

What kind of bracelet, where is it, do they have parts to repair - was a manual look up

Use AI to take picture of product - to find it in the catalog, single view of product from app



Source: SAP

Enterprise AI is a holistic approach to machine learning

Integration into production application

Take AI and bring it into the enterprise to manage and scale it

Make sure data is cleansed

Not lift and shift

Ensure it is quality data

Manage the development - Jupyter notebooks, use algorithms they want, not have it isolated

Manage the delivery - getting it to production

Scale AI across the enterprise



Source: SAP

Intelligent Information Management

Data Intelligence - evolution of Leonardo ML + Data Hub

Above is a framework for AI



Source: SAP

What is Data Intelligence

Described in the middle of above

Everything from connecting, governing data, preparing data, Jupyter notebooks, R, python, deploy, integrate, monitor and scale

A data science platform, to harness and scale data science platform



Source: SAP

Marketecture picture

Intelligent Information Management is foundation, connect to any data source

Pre trained models, and bring your own custom models



Source: SAP

Creating the AI assembly line

What does it take to make AI a repeatable process

Start with prepare models

Apply machine learning

Deploy - scale models

Automate business processes



Source: SAP

Goes in to detail on subsequent slides



Source: SAP

You can use the tools you know

Spin up any lab as a service

Deploy to production easily



Source: SAP

Understand and explain - track model performance; explain impact, results from algorithm and why



Source: SAP

Automate the low value tasks

Provide single interface

Connect to data, for governance

Run on the cloud, serverless cloud, only pay for what you use (SAP Cloud Platform)

 

Link to the replay is here

What do you think?