At a recent gathering of startup founders at his home in San Francisco, Larry Ellison, Oracle’s indefatigable executive chairman, CTO, and cofounder, had some fun with the wacky notion that the cloud platform, service, and infrastructure giant had somehow become a “dinosaur.”
Ellison told his guests that making Oracle ever-more successful and proving the naysayers wrong are always top of mind, joking that he thought about bidding on the giant Tyrannosaurus Rex fossil exhibit in the American Museum of Natural History and displaying it in the Oracle lobby. “Yeah, a dinosaur. But a very dangerous dinosaur,” he said.
At Oracle OpenWorld this month, Oracle executives, partners, and customers revealed just how dangerous an innovator the company continues to be. They laid out cutting-edge products and services powered by a range of emerging technologies, including machine learning, predictive analytics, blockchain, and the Internet of Things. What follows are just six highlights.
Machine Learning Makes Sense of the “Weird”
Calling machine learning a “technology that’s every bit as revolutionary as the internet,” Ellison described how Oracle is building machine learning-based databases and applications that get smarter as they take in more and more data.
Machine learning is based on algorithms that can learn from data without rules-based programming. “The more data we have these computer systems look at, we say we’re ‘training’ the computer system,” Ellison said during his second keynote at Oracle OpenWorld, on October 3. “And as the computers begin to identify patterns in the data, identify abnormalities in the data, they then can make predictions. And as we give the computers more and more data to train on, we say the computers are learning if, and only if, their predictions get more accurate. That’s the test. We give it more training data, their predictions get more accurate.”
Ellison offered the example of how machine learning algorithms will help companies improve their information security by analyzing reams of logged data and flagging anomalous patterns before intruders can do damage. He cited a fictional California-based retailer being able to proactively block someone posing as the CFO from logging into its finance systems from a computer in Ukraine. “Either she’s taking a really weird holiday in Eastern Europe, or this is a problem,” Ellison said.
- Related: Experience Oracle Cloud for free
The good news: Machine learning requires massive amounts of data to understand normal patterns and then detect anomalies, and Oracle produces “billions and billions” of log records in its cloud data centers, he said. “And that data can be used to train computers to say, ‘This is normal. Ignore 99.9999% of the data. But logging in from the Ukraine on a military base—that’s not normal. Certain kinds of SQL queries from that user—that’s not normal.’”
Say What, Chatbot?
Chatbots are one example of the entirely new ways people will interact with applications, going well beyond the web browser and smartphone, said Thomas Kurian, Oracle president of product development, during his Oracle OpenWorld keynote.
With chatbots, a customer can send a text message on any number of messaging platforms—Facebook Messenger, Slack, Skype, various voice-powered assistants—to ask questions such as: “How much is my electrical bill this month?” and “What tickets are available for tonight’s concert?”
The bot-building capability in Oracle Mobile Cloud Service lets companies create these applications. It provides the natural language processing platform so that a bot can understand the various ways people might ask a question and respond in a conversational way. It also includes a machine learning capability so that the bot gets continuously better at understanding and replying to questions. And it has integration capabilities so that the bot can access backend enterprise systems to find answers.
Such bots can also help employees, letting them use text or voice to ask ad hoc, conversational questions about production levels or a marketing campaign’s results. Kurian demonstrated an Oracle Analytics Cloud capability that sends data analysis as a feed to a bot on a mobile phone. Or a bot could text or call a manager automatically when data indicates a problem that needs attention.
Kurian described other ways to interact with applications, such as taking a picture and having an enterprise system recognize and react to it. For example, a consumer could send an ecommerce site a Twitter image of shoes she likes and ask if the site has any in stock.
“Our vision for the human interface for applications is [for it] to become seamless for humans,” Kurian said. “No longer is it just web and mobile screens, but you could speak to the application. You can interact with it with messaging. You can take pictures, and we can identify images, compare them with other things, and automate transactions.”
How the “People People” Can Also Become Data People
Employees come and go, but losing top performers can cost companies up to two times their annual salaries. Data analytics and machine learning promise to save employers many millions of dollars each year by helping HR managers identify the highest performers and understand what motivates them to stay or leave.
At an Oracle OpenWorld session, Charlie Berger, Oracle senior director of advanced analytics and machine learning, showed an audience of HR managers how they can use the analytics available in Oracle HCM Cloud. Specifically, Berger demonstrated how they can run attrition analysis against a number of pre-populated variables—such as the number of years an employee served in her current position, how many sick days she took in a given year, and the date of her last salary change—to determine possible linkages to her recent resignation.
Berger also demonstrated how HR managers can document the specific reasons employees leave their companies, and then feed that information into the Oracle HCM Cloud system to do predictive analysis to head off future departures. “The time it takes to build predictive models and score employee behaviors can be done in a few seconds, down from several weeks,” Berger said.
Blockchain Is the New Facilitator
Blockchain is one of the most hyped technologies of our time, and for good reason: It promises to lower the costs associated with many types of transactions, while speeding their execution and making them more secure.
The blockchain data structure lets multiple entities share data on a universal ledger without a central authority. Each entity controls its assets using a private key and independently verifies all transactions. Transactions added to any block in the chain are validated by multiple entities participating in that chain. “Blockchain can remove the need for intermediaries and replace it with cryptographically secure protocols,” says Mark Rakhmilevich, Oracle senior director of product management and strategy.
Oracle Blockchain Cloud Service provides preassembled blockchain code optimized for standard business processes, including ERP transactions that have traditionally required third-party validation. Blockchain addresses the problem of trust between organizations by providing independent validation through a tamper-resistant, peer-distributed ledger, thus eliminating the need for offline reconciliation.
Blockchain can manage common transactions, such as payment records, safety inspections, building permits, mortgage and loan records, purchase orders, and invoices. Members of a given supply chain could also use blockchain. For example, a food retailer could use it to check the provenance of its meats (and even see government inspection certificates). And the retailer, meat processor, and original seller could exchange fiduciary documents, validate shipments, and reconcile orders and invoices, all without needing a bank or other third party to validate those transactions.
Oracle provides its cloud service as a managed platform as a service, so that customers “don’t have to stand up new instances for every use case,” Rakhmilevich notes. In this preassembled service, “all needed components have been provisioned,” he says.
Developers Tap the IoK—Internet of Kegs
When it’s time to demo the latest application of the Internet of Things for a bunch of developers, you might as well have fun. Serving them beer is also a plus.
In a live demo at the JavaOne conference in San Francisco last week, members of the Oracle Developer Community, working with a local brewery, Alpha Acid, showed how an IoT setup could improve the beer manufacturing and product-improvement processes.
Mark Vilrokx and fellow members of the emerging technologies team in Oracle’s AppsLab group worked with Alpha Acid to install sensors in the brewery to collect data on various facets of beer making, such as temperature, pressure, and CO2 levels in the fermentation tank. “Before, if the brewer wants to know how warm the brew is, he pours it on his hand,” Vilrokx told visiting developers. “That’s not very scientific. And he gets a measurement for now, but maybe 10 minutes later it has cooled down and he doesn’t know.”
Vilrokx and his team tracked measurements over time and provided a chart where “hopefully it’s always within the range and you can see it,” he said.
The next part of the product-improvement process came when tasters (in this case developers at JavaOne) rated different batches of beer. Using the sensor data collected at the Alpha Acid site, the brewer can correlate the relevant beer-making data with the feedback on the batches that were rated highly and not so highly, allowing him to improve his processes over time, Vilrokx explained.
The demo in the JavaOne Developer Lounge was also outfitted with sensors linked to Oracle Internet of Things Cloud Service, letting Vilrokx’s team measure the number of pours, the number of ounces in each pour, and the amount of beer left in each keg, all visualized on tablet screens at each keg.
Emerging Tech Needn’t Make IT Jobs Obsolete
What do Oracle database administrators and other IT pros do when they’re experts in legacy systems but their employers are eager to embrace emerging technologies such as artificial intelligence and the Internet of Things?
“You have to tie what you do as a technologist back to value for the business,” advises Ross Smith, chief architect with Oracle integrator PITSS America. “The legacy systems that our companies want to modernize are often crucial, so our job as the data experts is to help the company express what they do in a new technology that’s forward-looking.”
So when the company wants to develop, say, AI-powered chatbots or industrial IoT applications, Smith says, “you have an opportunity to talk to your applications teams and say, ‘We’ve got new features in Oracle, we’ve got these cloud services, and we’ve got alternatives like NoSQL. We’ve got plenty of things that we can try out.’ So now you are suddenly recommending forward-looking solutions to a business problem. It’s our job as data experts to bring those opportunities to the floor.”
Rob Preston is editorial director in Oracle’s Content Central organization. Content Central staff members Jeff Erickson, Sasha Banks-Louie, Chris Murphy, and Michael Hickins contributed to this article.
