|
|
|
|
AI Cuts IT Problems at Fannie Mae; the Rising Risk of ‘Shadow AI’; Fujitsu Develops Emotion AI Tech
|
|
|
|
|
|
|
|
|
Fannie Mae’s headquarters building in Washington. PHOTO: ANDREW HARRER/BLOOMBERG NEWS
|
|
|
|
AI cuts IT problems at Fannie Mae by a third. An artificial-intelligence operations tool used by Fannie Mae has cut by a third the number of monthly issues its information-technology staff needs to handle, and more reductions are expected, reports WSJ Pro’s Jared Council.
Using machine learning to analyze tech problems. The mortgage-finance company began implementing the tool, developed by Moogsoft Inc., a year ago. The “artificial intelligence for IT operations” system, or AIOps for short, uses machine learning to analyze technical problems by tracking patterns and anomalies, isolating the causes of crashes or other malfunctions and suggesting a course of action—resulting in faster systems restoration.
Before the AIOps tool went live, Fannie Mae said its IT operations team was handling about 20,000 monthly incidents, such as a storage or application failures. But some of the alerts for those incidents were redundant or had the same probable root cause. The AIOps system learns how to correlate relevant alerts together and highlight the likely underlying issue.
[continues below]
|
|
|
|
|
|
|
Reducing alerts, speeding fixes. Fannie Mae’s 20 information-technology operations teams are divvied up based on business units. The first eight teams to receive the AIOps tool have seen a 35% reduction in incidents over the past 12 months. Fannie expects that when it deploys the AI system to all business units and the system gets better at pinpointing root causes, monthly incidents will decline by 50% to 60% from that 20,000 figure over the next year.
The AIOps system allows the staff to spend less time sifting through alert noise and more time collaborating to solve important tech issues, said Jay Rudrachar, Fannie Mae’s director of enterprise monitoring, analytics and reporting. This allows the teams to handle issues faster.
The teams using the AIOps tool have cut the time needed to resolve problems by between 25% and 75%, depending on the issue, Mr. Rudrachar said. Some problems that used to take several hours to fix are now handled in a matter of minutes, he added.
|
|
|
|
|
Organizations see risk in ‘shadow AI.’ By 2022, Gartner says 30% of organizations using AI for decision-making will come up against “shadow AI” as the biggest risk to effective and ethical decisions, reports the WSJ’s Sara Castellanos. Shadow AI refers to data outside the ownership of the IT department that is used to build AI models.
The concept is gaining popularity as more AI tools become available to employees without formal training in data science or artificial intelligence, according to Gartner. “While this can be a productivity bonanza...there is also a challenge with unleashing these powerful tools on an untrained audience,” according to a Gartner paper on the Top 10 Strategic Technology Trends for 2020.
|
|
|
|
|
|
|
|
|
|
Freddie Mac named Frank Nazzaro executive vice president and chief information officer. PHOTO: FREDDIE MAC
|
|
|
|
Freddie Mac CIO looks to modernize IT. Freddie Mac’s acting chief information officer, Frank Nazzaro, is taking on the role permanently as the mortgage-finance company looks to digital technology to increase loan approvals for a wider group of qualified home buyers, reports WSJ’s Angus Loten.
Mr. Nazzaro has served as acting CIO since May, taking over from Stacey Goodman, who became CIO of Prudential Financial Inc. Reporting to Freddie Mac Chief Executive David Brickman, Mr. Nazzaro will lead the information-technology division while setting an overall technology strategy.
The challenge for the mortgage industry is the need to balance aggressive IT modernization efforts with minimizing the impact to customers and end users, Mr. Nazzaro said. “The overarching goal of this work is to improve velocity, reduce cost and increase quality throughout the mortgage process,” he said.
Freddie's digital initiatives. He said Freddie has a range of digital initiatives under way, including the use of advanced analytics, machine learning and other emerging tools aimed at making processes more efficient and better managing risk.
|
|
|
|
|
|
|
|
A U.S. Customs and Border Protection officer inspects a particular vehicle on the international border bridge Paso del Norte, as seen from Ciudad Juarez, Mexico, in July 2019. PHOTO: DANIEL BECERRIL/REUTERS
|
|
|
|
Border patrol seeking information on facial recognition for body cams. The U.S. Customs and Border Protection last week put out a request for information on body cameras, cloud storage and video management software that could assist agents checking for contraband and for immigrants illegally entering the country, reports Reuters. CBP listed a number of features, such as encryption, as critical. It put in facial recognition as a non-required item of “potential interest.”
|
|
|
|
|
|
|
Fujitsu develops new emotion AI technique. Fujitsu’s laboratories develop AI-based technology that can pick up slight changes in a person's facial expressions, such as nervousness or confusion, reports ZDNet. While other emotion AI tools can recognize expressions, they are limited to eight “core” states: anger, contempt, fear, disgust, happiness, sadness, surprise or neutral, according to the report.
Process requires less data. Existing systems need a lot of data to recognize facial muscle movement, or action units (AU). Fujitsu said it achieve its results by a process that extract more data from one image. It turns pictures taken from an angle into an image the looks like a frontal shot. The new image lets the AI detect AUs much more easily and more accurately, according to the report.
“With the same limited data set, we can better detect more AUs, even in pictures taken from an oblique angle, for example,” said a Fujitsu spokesperson. “And with more AUs, we can identify complex emotions, which are more subtle than the core expressions currently analyzed.”
|
|
|
|
|
|
The Google Pixel 4's orange color is hard to miss. PHOTO: KENNY WASSUS/THE WALL STREET JOURNAL
|
|
|
|
Google shows off AI prowess with smartphone. Google’s Pixel 4 might be the most advanced phone you can buy, says the WSJ’s Joanna Stern in a review of the device. Artificial-intelligence software allows for stunning nighttime photos, better and faster communication with Google Assistant and the ability to process speech-to-text in real time without an internet connection.
|
|
|
|
|
Media-monitoring startup raises $25 million. London-based Signal AI will use the money raised in its Series C round to develop software that scours news sources and other public data to look for trends and risks, TechCrunch reports. Redline Capital led the round, with GMG Ventures, a unit of Guardian Media Group, and fellow media-industry investor Hearst Ventures also taking part. The valuation wasn't disclosed.
|
|
|
|
|
Indiana University to build AI center. Fred Luddy, an Indiana University alumnus and the founder of ServiceNow, a Silicon Valley-based IT company, donated $60 million to IU to build the Luddy Center for Artificial Intelligence, according to a report on IndyStar.com. The center will fall under the university's School of Informatics, Computing and Engineering in Bloomington.
|
|
|
|
|
|