South Park Kids Make Jokes and IoT Devices Seriously React In the debut episode for South Park’s 21st season, the beloved characters not only poked fun at the white nationalist movement in an episode called 'White People Renovating Houses', but also yelled commands at their own cartoon Alexa devices throughout the episode. This actually started activating Alexa and Google Home models in the homes of viewers. In the episode, after proclaiming that smart devices are stupid, South Park character Eric Cartman got a smartphone and an Amazon Echo speaker. Cartman’s interaction with Alexa starts innocently enough, with a request to Alexa to set an alarm and tell a joke. The best moments definitely came as the South Park kids asked Amazon Echo’s Alexa to do and say increasingly disgusting things. “It’s a very dumb joke that never gets old because who hasn’t asked their smart home device something idiotic?” This has been seen before. Earlier this year, in Burger King’s “Whopper Burger” television ad, they  activated Google Home devices and had them recite the Wikipedia entry related to Burger King’s Whopper. “The reaction at the time ranged from laughter to outrage. In the end, Burger King and its ad agency won a top industry award, a Cannes Lions for the stunt.” Stirring Up Trouble with IoT Devices Leads to Serious Reactions on Twitter Not only was it revealed that there are people who own both an Alexa and a Google Home, but more seriously that the show could stir up trouble for both. As viewers got further into the episode, some viewers even had to unplug the listening mechanism on their devices. Viewers took to Twitter to share videos and comment on what the episode was doing to their own devices. This episode of South Park drove my Alexa crazy. I never knew my Alexa had such a potty mouth. pic.twitter.com/lWIHySClRd — Tony French (@TonyLFrench) September 14, 2017 New South Park episode is making Alexa go nuts. pic.twitter.com/Rs8oNL7s2u — Tom Buros (@TomBuros) September 14, 2017   Breaking Boundaries Exposes an IoT Weakness Not only was this intrusion a massive violation of entertainment’s fourth wall,  but there is a much greater issue here with the appearance of a new form of ‘digital voice attack’, further weakening our boundaries. “The fact that Alexa doesn’t identify users means that for “her,” all us humans are the same. Actually, you don’t even need a real person to speak.” The world of Internet of Things (IoT) introduces new types problems. While we have felt a general sense of security and control, this act and the targeted technology forces us to redefine our boundaries - borders that are much harder to define and measure because they are not definite. Detecting When Something Goes Wrong with Anodot AI Analytics Anodot actually already warned about the potential for this happening back in February. This raises the underlying question: How do we detect and recognize when these now ubiquitous AI-controlled systems get something wrong? Most AI systems aren’t quite capable of dealing with these difficult, anomalous behaviors. More importantly, this situation could have been avoided, but should serve as a warning to companies making real-time business decisions based on streaming data—there needs to be systems in place to detect when something out of place happens, and to let the right people know. Anodot would have flagged the dramatic spike in search terms and alerted the responsible Amazon team to the issue to investigate further.  Anodot would have been crunching Amazon’s time series data to determine the normal range for these searches. When the spike in Alexa-powered shopping lists rose beyond normal limits, Anodot would have triggered an action and applied a significance score that could help Amazon determine how fast and comprehensive the response might have been.  

Kickstart your business monitoring - see how machine learning anomaly detection can provide your team with the kind of spot-on, real-time alerts that prevent costly incidents and protect revenue.

Business KPI dashboards don’t provide actionable insights now Business data analysts are in dashboard hell right now. They have to interpret data from so many different sources and then try to figure out what is the best action to take, with the most important decisions being made based on this information.  Despite having “Key Performance Indicators (KPI) Dashboards” for business, they struggle to get an integrated view of all business metrics. With greater volumes of data being collected, data analysts can’t keep up with the pace. Companies are making tremendous investments in dashboard and reporting technologies, like Business KPI dashboards, to keep better tabs on their operations. You’ve probably seen business dashboards with multiple KPIs that show sales figures, customer satisfaction score, churn etc. While designed to provide high level visibility of different business metrics, nevertheless data analysts simply can’t keep up with the demand to crunch all the data and then extract answers from their business KPI dashboards. Business KPI dashboards provide too much information and too little insight There are many reasons why business KPI dashboards can’t be relied on to provide you with actionable insights in real time. While an analyst can analyze data based on one or two dimensions (for example, device and geo), when an issue happens, they then have to manually add more KPI dimensions, spending a lot of time to find the source of an issue. Standard business KPI dashboards fall short when it comes to usability, specifically the accessibility of the important signals in the data.  Unable to automatically highlight what’s important, dashboards require analysts to process the data manually and iteratively, and still lack the ability to instantly drill down into granular metric-level data.  Traditional analytics and BI solutions, like business KPI dashboards, deal with historical data, not this minute, not showing a real-time status. Due to their limitations, business KPI dashboards typically look at only a subset of all the available data. These limitations yield at best delayed and at worst incomplete results. Business KPI dashboards simply cannot provide intelligent correlation The usual business KPI dashboards can’t understand a KPI in the context of the many, complicated, direct and indirect relationships between all your metrics. Data complexity, data-type growth, and data volumes threaten to overwhelm the interface, weakening the dashboard’s consumability.  Locating information takes profound familiarity of each dashboard metric, adequately see events and an elephant’s memory as you to try to visually correlate the data from across the organization, trying to figure out what to do next. When a BI team has trouble keeping up, they often get unpleasant surprises, discovering issues and opportunities long after the financial or reputation damage is done. Despite configuring multiple business KPI dashboards, they can still face many data blind spots. With data constantly changing, looking for answers on business KPI dashboards is a struggle, as something else could come in from left field, totally off the radar, affecting the results. AI analytics identifies trends and provides correlations in real-time Instead of long investigations and analysis through multiple business KPI dashboards and making manual correlations, business analysts can rely on AI analytics to probe deeper into the data and correlate simultaneous anomalies, revealing critical insights into operations. A real-time, large-scale automated anomaly detection system using machine learning methods can free data analysts from constant manual monitoring around just a few KPIs. When working with thousands or millions of metrics, you can’t just hire a staff of thousands of analysts to analyze your data for key decisions. Using automated significance ranking of detected anomalies, data analysts can focus in on the most important business incidents. This level of automation can provide actionable insights in real-time. Insights built upon deep learning, especially in complex, fast moving digital industries like ad tech, add another level of protection to company revenue. Both Microsoft and Google rely on advances in deep learning to increase their revenue from serving ads. Advanced machine-learning analytics allow ad tech companies to identify trends and correlations in real time, like instantly correlating a drop in a customer’s bidding activity to server latency. An automated AI analytics solution maps out the relationships between all your metrics, even if they number in the millions. Turning on all of the lights in the room, instead of using just a flashlight, allows companies to correlate all relevant data, not just the typical collected information presented in business KPI dashboards. By taking in the full picture, you can make better decisions to impact business success and improve customer satisfaction. And that’s where AI analytics really shines. Unlike traditional BI tools, by detecting the business incidents that matter and identifying why they happen, AI analytics lets you remedy urgent problems faster and capture opportunities sooner. Our automated AI analytics solution uses the latest breakthroughs from machine learning and data science to give our customers actionable insights in real time, something they were not able to achieve from their dashboard based business intelligence, bringing valuable business opportunities and insights to the surface.  

How does eToro stay on top with so much changing and vital data to be aware of? We visited their headquarters and met with Elad Gotfrid, eToro’s Director of IT, who shared their approach for handling the company’s data.

Learn how leading eCommerce and retailers are leveraging the power of machine learning - identifying problems and leveraging business opportunities faster.

GetYourGuide's engineering team leverages anomaly detection across the business, from monitoring revenue to watching for brand hijacking and affiliate fraud. Anodot alerts bring attention to the underlying issues much more quickly than any other method, which is helping prevent financial loss and keeping their global business running smoothly.

KPIs are the Cliff’s Notes of business metrics: a handpicked selection of the many measurable quantities which a business can - or does - collect. They can provide vital clues as to how well the business is doing as a whole, for instance which marketing initiatives are and aren’t working, and thus they provide feedback and can guide decisions on how resources are expended, for instance discontinuing a promotional offer that’s not driving any new sales. KPIs are considered “key” often because managers deem them directly related to profit, which is why metrics such as year over year same-store sales are such a common KPI. In order to tell a clear and convincing story, a given KPI needs to be transformed by analysis and visualization from raw data into business insight, and the correct way to do so isn’t always obvious. This is why traditional KPI analysis requires human data science knowledge and is very iterative, two aspects which prevent discovering actionable insights in real time. Traditional KPI Analysis In order to understand how Anodot’s full-stack solution makes dashboard-based KPI analysis tools obsolete, it’s helpful to review the typical KPI workflow which relies on those tools. First, the underlying data for a KPI needs to be pulled in from whatever files or databases they live in. Then, that data needs to be structured or formatted in a way that makes analysis smooth. This step is called “data cleaning” and is often required if the files containing the data for the KPI have rows that serve as table titles or headers, or if data from two sources needs to be combined before analysis. Finally, the data is imported into the dashboarding tool, where it is analyzed and visualized. Usually, the analysis and visualization isn’t a single step, but rather multiple iterations of reviewing, giving feedback and implementing changes. This is due to the fact that a whole team of analysts and data scientists often work on a single KPI, and the multiple iterations are needed before everyone on the team is confident that the visualization of the analyzed data will meet the requirements of management. Although there are obvious benefits to this collaborative approach, it is very time-consuming and takes data scientists and analysts - of which companies have a limited supply - off of other tasks. Furthermore, even after a KPI visualization is created and added to a dashboard, it may be modified or even replaced by another one when it is reevaluated as time goes on and business goals are (or aren’t) met. Once created, the dashboard visualization of the KPI is then monitored either manually or semi-manually via the use of static thresholds and alerts. Drawbacks to traditional KPI analysis, tracking & monitoring We’ve already mentioned the problems of tying up limited data science and analyst personnel as well as the fact that KPI analysis via dashboards is far too time-intensive to alert you to business incidents in real-time, but there are other drawbacks to this approach as well. Since KPIs tend to be aggregated metrics (like averages or totals), they can hide isolated yet significant problems or opportunities, as we've seen in temporary ecommerce glitches, such as listing a top-brand headset as free, rather than at its normal price, even for just a number of hours. Also, this KPI analysis workflow can very easily miss important signals in your data if you choose the wrong KPIs for monitoring to begin with. Lastly, the low number and aggregate nature of KPIs results in data that isn’t very granular. Granularity, as discussed in our previous post, is required to actually understand and fix the specific problems discovered on the dashboard. Monitoring your KPIs with Anodot: a comprehensive tracking solution for all your metrics As we mentioned above, Anodot provides a “full-stack” real time AI analytics solution. Not only can it accurately detect anomalies in all your metrics in real-time - eliminating the need to select a handful of KPIs - but Anodot’s correlation of related anomalies gives your analysts the benefits of granularity (specific actionable insights) with the conciseness you’d expect from a traditional dashboard tool. By giving you built-in data science, the majority of manual KPI analysis can be replaced by intelligently correlated and collated anomalies which point directly at business incidents. Anodot is scalable to millions of metrics, far beyond the manual, iterative approach that relies on traditional BI dashboard tools. That scalability enables Anodot to track all your metrics, thus giving you the granularity you need to find the important signals in all of your data, not just the specific KPIs you think you should be looking at. The more pixels, the sharper the image, and the key advantage of AI analytics is the ability to learn both what’s important and what’s related, not only showing you what’s there in the image, but also what it means.

The Production Services team at leading fintech Payoneer uses Anodot Autonomous Detection integrated with the Coralogix Log Analysis Platform to accurately identify system issues with real-time alerts and to accelerate remediation time.

Recently Gartner released their latest Hype Cycle report for Data Science and Machine Learning, advising Data and analytics leaders to use this report to better understand the data science and machine learning landscape and evaluate their current capabilities and technology adoption prospects. The report states that "to succeed with data and analytics initiatives, enterprises must develop a holistic view of critical technology capabilities." One of the technologies noted that is 'On the Rise' side of the curve is Advanced Anomaly Detection. Advanced Anomaly Detection is Accelerating Expecting Advanced Anomaly Detection to accelerate to mainstream over the next ten years, the Gartner report explained that "Anomaly detection is relevant in situations where the existence, nature and extent of disruptions cannot be predicted. Systems that use advanced anomaly detection are more effective than those that use simpler techniques. They can detect subtle anomalies that might otherwise escape notice, provide earlier warning of impending problems or more time to capitalize on emerging opportunities, reduce the human effort required to develop a monitoring or measuring application, and reduce the time to solution for implementing complicated anomaly detection systems." In the report, Gartner senior analyst Peter Krensky and Gartner VP and distinguished analyst W. Roy Schulte provide their insight for how Advanced Anomaly Detection will impact companies, and the benefits to be gained. They say that “virtually every company has some aspects of operations for which it is important to distinguish between routine conditions and matters that require extra attention. Much of advanced anomaly detection is a competitive advantage today, but we expect most of the current technology driving it will be widespread and even taken for granted within 10 years because of its broad applicability and benefits. Advanced anomaly detection is already being embedded in many enterprise applications.” Forces Driving the Need for Advanced Anomaly Detection We find the main drivers that are creating the need for Advanced Anomaly Detection solutions come as businesses are having to manage increasing amounts of data and tracking more metrics. An unavoidable consequence of the scale and speed of today’s business, especially when lines of code sets are changed in seconds, are costly glitches. Moreover, as many processes happen simultaneously, many businesses monitor activities by a different person or team. Anomalies in one area often affect performance in other areas, but it is difficult for the association to be made when departments operate independently of one another. With businesses constantly generating new data, they need a solution that can analyze metrics on a holistic level and give people the insights they need to know what is going on. Applying Advanced Anomaly Detection Looking at how Advanced Anomaly Detection has already been applied, the analysts note that "its use is increasing in applications as diverse as enterprise security, unified monitoring and log analysis, application and network performance monitoring, business activity monitoring (including business process monitoring), Internet of Things predictive equipment maintenance, supply chain management, and corporate performance." High Business Impact of Advanced Anomaly Detection Advanced Anomaly Detection received a maturity level of ‘Emerging’, meaning that it is expected to grow and evolve. The strong momentum that Advanced Anomaly Detection has gained has drawn the attention of innovative vendors. The report lists some sample vendors who offer Advanced Anomaly Detection solutions, such as Anodot. Our Take on the Gartner Hype Cycle on Data Science and Machine Learning Each year, Gartner produces Hype Cycle reports to provide a graphic representation of the maturity and adoption of technologies and applications, and how they are potentially relevant to solving real business problems and exploit new opportunities. In our case, the challenge with big data is how to find the right insight, or really what are the right business incidents in the data. While Gartner outlines a clear need and drive for Advanced Anomaly Detection for data driven businesses, we believe that mainstream adoption will appear faster than projected in the report, because businesses that implement it will enjoy clear competitive advantage, driving competitors in their segment to move quickly to add Advanced Anomaly Detection capabilities. There are many systems that try to find events and data that is not normal. Yet, often these systems fail, identifying too many anomalies (false positives) or not enough (false negatives). Applying machine learning to many rapidly changing environments means constantly updating and retraining models, to prevent false positives. Our AI analytics solution offers autonomous analytics. Instead of having to ask the many questions and carry out complex data analysis, Anodot does the work, and provides the answers to understand why an incident happened. Next Step Click here to download our white paper that helps you explore and weigh the issues around the Build or Buy Dilemma for Advanced Anomaly Detection. All statements in this report attributable to Gartner represent Anodot’s interpretation of data, research opinion or viewpoints published as part of a syndicated subscription service by Gartner, Inc., and have not been reviewed by Gartner. Each Gartner publication speaks as of its original publication date (and not as of the date of this report). The opinions expressed in Gartner publications are not representations of fact, and are subject to change without notice. Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. Gartner, Hype Cycle for Data Science and Machine Learning, 2017, 28 July 2017