Want to be more productive? Boost ML usage with Geneos 7 and Opsview Enterprise
Artificial Intelligence (AI) is a massively loaded term that has been around for many years and encompasses a large range of study whereby humans work to make machines exhibit intelligence. Intelligence can be subjective, of course, but our Geneos and Opsview power users are looking for tools that act as companions to help them get ahead of issues and better characterize the structure and operation of their critical applications.
More recently, AI has presented itself as a more human companion through chatbots, search engines, Generative AI (GenAI), and other forms, but ITRS' incarnation doesn't have this human element.
To take on such a broad subject as AI, ITRS needed to pick a focus area. That area was a place that would result in the optimization of new and existing deployments of ITRS software. Specifically, the Anomaly Detection part of Machine Learning (ML).
This is where Obcervant comes in. Obcervant was developed to be our umbrella Obcerv application covering all the features and functions that would be the companion making customers' operations teams more efficient. Starting with the Forecaster, we aimed to take the overhead out of using a supervised learning technique to manage the growing list of metrics being collected. Recently, we announced the Dynamic Threshold APIs that allow our tools to request new thresholds for current times and, without user intervention, change the thresholds in our real-time monitoring calculations.
Our goal is to make operations teams more efficient. We want to enable them to spend less time administering the monitoring system for coverage and more time exploring the coverage, mining data, and ultimately finding issues before they lead to business outages.
Rapidly changing environments, faster development deliveries, and growing transaction volumes have made these functions more critical than ever.
Imagine monitoring your 10 critical company websites used by internal and external teams. Your application is global, so you need to verify connectivity from 15 different locations including designated countries and office locations. Each of those websites is collecting three KPIs on user response time. Finally, let's assume you have configured checks for three different versions of your software.
This isn't just a single rule anymore. This isn't just a single threshold. Multiply all that together and you're now managing 1,350 metrics each with a different Normal. You would expect that your website in New York serves local users faster than users on a different continent or a user connecting over a slower VPN connection. Using that data, our Machine Learning capabilities will determine and apply the right thresholds, so you only get notified when something is deviating from normal. That full circle feedback loop is exactly the definition of Machine Learning.
Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data, and thus perform tasks without explicit instructions. (source)
There is no one algorithm to rule them all, so ITRS is continuing to develop Obcervant components to collaborate with our clients' operations teams and monitor the most common types of metrics our customers collect: latency, capacity, seasonal, status and log.
To explore how ITRS can help you preemptively avert issues and drive operational efficiencies, get in touch with our solution specialists today.