Today, Oracle Openword kicked off in San Francisco, CA. During one of the first sessions of OpenWorld, Amit Zavery (@azavery), Executive Vice President of Fusion Middleware and PaaS Development, walked us through where Oracle is going with all of their cloud platforms but with a special focus on PaaS.
Learn about the strategy and vision for Oracle’s comprehensive PaaS solutions. See demonstrations of some of the new capabilities built into Oracle Cloud Platform including a trust fabric and data science platform. Hear how Oracle’s application development, integration, systems management, and security solutions leverage artificial intelligence to drive cost savings and operational efficiency for hybrid and multi-cloud ecosystems.
Oracle started investing in the cloud about 7 years ago and now had a range of PaaS services:
- Integrate and extend
- Publish and engage
- Analyze and predict
- All cloud components use the same data model, are integrated, and use a common security model
- has AI built-in
- It’s very scalable and highly performant
- Full blown AI
- Digital Assistants
- Virtual and augmented reality
Oracle Cloud is open. it’s built to be open from the ground up.
- Support most programming languages in
- Databases like Oracle, Caasandra, Mongo, and MySQL
- Range of tools and frameworks
- Kubernetes, Kafka
- Hyperledger fabric
- AI with Tensor Flow,etc.
- Heterogeneous with support for
- Amazon Alexa
- Security includes OAuth and other key investment
Basically, Oracle takes the open source and deploys/supports it. It’s not a fork. There is no lock-in.
Fact: Oracle has 18,000 customers
Fact: Oracle has $26 Billion in revenue for cloud in FY 18
Analyze and Predict
There is a range of challenges. The tools are fragmented, the data is not well governed, and it’s labor intensive to train an AI. To make it easier, Oracle has been creating a platform to address these challenges.
It’s a single integrated end to end solution.
Machine Learning is used to enhance user productivity
Rary Beauty needs to expand their retail presence. Jax showed some capabilities
- Started with a conversation in slack with Greg Wallis, an operations analyst
- Greg wants to understand nearby skilled resources
- Greg invoked the Oracle Digital Assistant and brought a map of Northern CA and related analysis
- Rashana clicked on the analysis which took her to Oracle Analytic Cloud (OAC)
- Clicked on the additional data source
- Note: OAC connect to over 40 data sources,
- Connected to Taleo talent cloud to bring in that data
- Brought in a simple list of Candidates with some key data like zip and interview score
- Used the OAC interface to layer the candidates on the map with the other data
- Added another filter to get very specific job titles
- which highlighted the job roles needed for the new stores
- Rashana returned to Slack and posted a message to Greg that they had a few gaps with one specific role and no candidates in the right area
Publish and Engage
You want to be able to take the content and engage the users. This includes:
- Build new experiences across many channels quickly
- Govern the experience with up to date content
- Optimize the experiences using the right data to make the right decisions.
Oracle built an API first CX stack to deliver the content through any channel in a number of hours. This is a full set of capabilities.
Picking up where Greg left off int he last demo.
- Assistant create a new posting
- Married natural language processing with a smart content creator
- Think of it as having fun with word vector math
- Used a template to set the city. Images and content change as the search narrows
- Set a restriction of 20 miles from the Pleasanton location
- Laid in an image to the job description
- Submitted for review
- Which was approved
- Published to several channels
- Returned to Slack for a conversation with Greg
- Greg sees the post on LinkedIn
- Gets a report in slack on the number of applicants and the 10 candidates flagged for follow up
- Kicked off an interview with Alicia
- Used a digital assistant to schedule the interview
- The DA setup the time set a reminder and verified contact information
- Time passes and Alicia wins the job and she needs to onboard
- She downloads the Rayr mobile application
- Sets up her identity, password, picture, etc.
- Confirms all her personal information
- Picks the color of her new tablet
Develop and Deploy
You have to be able to create new applications. This includes building, monitoring, etc. It’s the whole develop and support lifecycle. Oracle has a complete portfolio of CICD tools to work together and deploy the apps.
Explore key considerations, integrating the cloud with legacy applications and challenges of current cloud implementations.
They also have a data science platform.
What it would take to find the best data science candidates and get them working
- Used a machine learning tool to understand who can be the best employees
- Launched a session
- Chose to compute resources to associate to the session
- Selected environments like Tensor flow of Sidekick learn
- Opened a session
- Had a range of tools
- Had tools to parse content like a resume
- Had the ability to use python based tools for coding, graphics, etc.
- As changes to the session were made, synced them back to github
- This produces a model which is ready for the development team
- Time to turn over to the Dev team
- Brough in the lead developer
- Published a REST endpoint so this model is available to push to Slack and other channels
- The endpoint came with wrappers for common languages and frameworks
Integrate and Extend
You have a number of challenges. The SaaS tools are decoupled. There’s no process automation. Oracle wants to solve these problems. They unified their platform. they have 100+ adapters to a range of internal and external applications. It’s hybrid and heterogeneous with built in intelligence.
Once that foundation is set, you need to add intelligence to the system.
There are two parts here. First is building the skills.
- A skill is a chatbot. The digital assistant is the point of interaction which may have many skills
- Part 1: define the dialog
- Showed an easy to use interface to set Bot message and possible responses.
- It’s a flow for the non-technical user
- Part 2: generate the flow
- Create all artifacts required by the engine to create the model
- Think of the phrases the end user could use
- Part 3: modify the flow with the different nuances
- Catchall situations
- Part 4: integrate into different channels
- Picked Trilio for SMS and set that up
- Voila: integrated to SMS
- Note: this is tightly integrated with integration recipes from the integration cloud
Part 2 of Demo the integration cloud
- Integrations are based on recipes
- Oracle HCM
- ERP cloud
- etc. (there’s a lot of them)
- Tools include a workflow for provisioning in the integration
- Announcement: Have a new RBA connector which lets you discover bots
- Activate the integration and start monitoring
Secure and Manage
- Using Oracle Public Cloud for Authorization
- Alicia took a photo of her credit card
- AI found and quarantined it as a violation
- A dashboard shows all the high-risk alerts
- Notes the origin, what policy was triggered and the action is taken
- Also shows that this is the first time Alicia has gone against policy
- Jump back to Alicia
- She’s back in the mobile app
- She has an alert with a one-time passcode from the Casby service
- Now she gets full access
- Note: this will not happen as Alicia shows more security responsibility