Safe Harbor Statement The following is intended to

Safe Harbor Statement The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle. Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 2

Data to Prediction - Dockerized Abhinav Shroff- @abhinavshroff Principal Product Manager Oracle Cloud Development Tools 4 th April, 2018 Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted

Machine Learning Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential

Apps Tons of Data Build Smart Systems Copyright © 2018, Oracle and/or its affiliates.

Data in Enterprise Generating Data Analyzing Data Pushing Feedback Copyright © 2018, Oracle and/or

Building Machine Learning Systems Problem Definition Consuming the ML service Preparing Data Identifying Algorithm Developing the ML Model ML model made available as a service Freezing the Model Validating the Model Capturing Feedback Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 7

Preparing Data Select Data • Data Availability • Data Filtering Preprocess Data • Formatting • Cleaning • Sampling Transform Data • Scaling • Decomposition • Aggregation Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 8

Machine Learning - Development Copyright © 2018, Oracle and/or its affiliates. All rights reserved.

Programming Language • Containerization support • Machine Language libraries • Domain for implementation • How you intend to Consume Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 10

Development Tool • Container friendly • Supports Collaboration • Multi Language Support • Standalone Snippet execution • Inline Charting Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 11

Machine Learning Model as a Microservice Jupyter Kernel Gateway • Extend ML model code as a REST service • Container based deployment • Consumable across all channels • Faster Development lifecycle • Efficient Dev. Ops Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 12

Deployment • Solves “works on my machine problem”. • Support for multi runtimes. • Eases out migration from dev to prod • Ideal for Microservice architecture • Ease of Clustering • Self Contained unit. Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 13

Container Orchestration Oracle Container Cloud Service Iaa. S Health Check Scaling Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 14

Machine Learning - Dev. Ops Copyright © 2018, Oracle and/or its affiliates. All rights

Dev. Ops Platform To stitch everything together Oracle Developer Cloud Service Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 16

Dev. Ops – Automation in Data Preparation Run Data Extraction Code Periodically Trigger Scripts to Format the Data Deploy DB Docker Image Create the DB Docker Image with Processed Data Trigger Scripts to Clean the Data Trigger Scripts to Aggregate / Decompose Data Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 17

Dev. Ops process for ML Systems Clone the Notebook Template from SCM Develop ML Model Code Push Notebook to SCM Collaborate and Develop Dockerfile Automated Application Test Deploy Image on Oracle Cloud Push Image to Docker Hub Build Docker Image along with Code Push Dockerfile to SCM Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 18

Machine Learning System- Architecture Copyright © 2018, Oracle and/or its affiliates. All rights reserved.

Reference Architecture – ML System on Cloud Saa. S as Data Source Data Preparation Automation Scripts ML Data Container ML REST Application Container Test Scripts Client Application to Capture feedback Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 20

Demo Usecase Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential

Demo Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential –

Summary • Use Containers and Cloud Dev. Ops platform for: Faster Development and Deployment Scaling fast Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | Confidential – Oracle Internal/Restricted/Highly Restricted 23

Try Oracle Developer Cloud Service Yourself cloud. oracle. com/developer_service • Tutorials • Videos • Whitepapers • Documentation • Forums Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | 24