The Next Generation of APM
Recently, several well-known research consultant firms have written about the emerging next generation of APM tools and capabilities. Gartner speaks of “…artificial intelligence for IT operations,” or AIOps. Enterprise Management Associates (EMA) refers to Advanced IT Analytics (AIA).
Regardless of terminology, this next phase of APM involves the application of AI-related technologies such as machine learning, predictive analytics and anomaly detection, data mining, streaming “fast data” analysis, and natural language search. Nastel AutoPilot’s key capabilities span all of this plus end-to-end transaction tracking across hybrid cloud, legacy, and mobile environments. See the Analytics page for in-depth architectural and feature descriptions.
The Next Generation of APM
Nastel’s AutoPilot next-generation APM analytics leaps ahead of other vendor offerings by:
- Amplifying the effectiveness and productivity of IT technology and business stakeholders
- Improving application performance, business efficiencies, and business outcomes—regardless of the technology environment.
By detecting problems and combining emerging trends and subtle behavior patterns into a clear picture of how IT operations impacts business outcomes, all stakeholders make more informed and rapid decisions based on business insights hidden in a multitude of high-volume data sources and tools.
According to EMA:
“Nastel’s AutoPilot is a distinctive example of application performance management (APM) combined with a core platform optimized for advanced IT analytics—separating it out from all, more standard, “APM” vendors. In addition, AutoPilot is uniquely directed at improving business efficiencies and business outcomes as they align with both endpoint and middleware-related transactions. This focus on business alignment and business outcomes also sets AutoPilot apart.”
Potential use cases include almost any industry vertical where, in addition to traditional APM monitoring requirements, subtle data outliers demand fast reactions and decisions. Some examples include actual or potential financial system security breaches, non-compliance, supply chain issues that potentially cascade rapidly into major delivery problems affecting customers…the list is almost endless.
- Nastel’s AutoPilot leverages traditional APM plus machine learning and other advanced capabilities to:
Continuously improve the detection of root causes of software problems in complex application ecosystems, before services are noticeably affected
- Reduce churn, support costs, and operational risk through improved productivity, faster IT problem resolution, and tool consolidation
- Relate the performance of intricate software systems and services operating across cloud, legacy, mobile, and IoT environments to business objectives
- Provide easy-to-use analytics tools for stakeholders like app developers, operations, DevOps teams, administrators, systems and infrastructure management, mobile and user experience specialists, and business people responsible for transaction-related business services and customer satisfaction
Representative Use Cases/Scenarios
DevOps – Because of AutoPilot’s distinctive combination of application and business insights, DevOps teams gain immediate benefits from the ability to identifying and monitoring interdependencies between development and operations for handoffs, feedback (including end-user experience), and minimizing developers’ time lost to troubleshooting. AutoPilot also can access GitHub repositories to support development more directly.
Business Impact and Business Alignment – Nastel’s maturity in business impact and business alignment sets it apart from most vendors. Not only does AutoPilot integrate and correlate observed transactions with business objectives—such as revenue, business activity metrics (BAM), business SLAs, business process impacts and user behavior metrics—its visualization and reports in mapping transactions to business outcomes are among the most mature and executive-friendly in the industry. Reports include financial payment topologies, order and shipment tracking, major compliance violations, and healthcare records claims processing, among many others. Its well-evolved capabilities for “sentiment classification” also set it apart in terms of business impact values.
Longer-Term Value: Toolset Consolidation – Nastel’s broad integration capabilities and technology-agnostic design philosophy provides a strong basis for extensive tool consolidation from an application and business service delivery perspective.
Unifying IT and Improving Operational Efficiencies – AutoPilot’s reach is sufficiently broad to provide an excellent top-down method for helping unify and optimize IT in a business context. Nastel also enables new areas such as social IT and improved mobile communication for IT stakeholders.
Integrated Security – Nastel AutoPilot monitors compliance-related behaviors for critical applications in finance, health, and other sectors. Moreover, Nastel can provide exceptional value in assessing fraud in context with transaction-related behaviors.
Change Impact Awareness – AutoPilot’s real-time awareness can immediately assess the impacts of changes, especially when affiliated with new application releases.
Capacity Optimization – AutoPilot’s insights into DevOps efficiencies, hybrid cloud efficiencies, and application optimization-related KPIs can provide valuable data to support capacity management and planning tools.
Technical Feature Breakout
Some of AutoPilot’s key features include support for both streaming analytics and historical data via its highly scalable Lambda architecture; machine learning for anomaly detection and behavior and sentiment classification; and end-to-end transaction tracking across hybrid cloud, legacy, and mobile environments. Some specific highlights are included below. For more details, see Analytics.
Nastel provides multiple machine learning methodologies that learn and improve their analysis over time without any dependency on writing rules. The methodologies include:
- Predictive Anomaly Detection
- Bayesian Conditional Probability
- Graph Analysis
- Root Cause Analysis
Predictive Anomaly Detection
Based on Robust Principal Component Analysis (RPCA), this algorithm:
- Perceives data outliers within massive amounts of event “noise” machine-learning-based anomaly detection
- Detects anomalies over a specified period of time
- Senses unusual elapsed times and learns if they become normal
Nastel’s Bayesian algorithm classifies behavior and learns over time. It can:
- Addresses the problem of judging documents as belonging to one category or the other (such as spam or legitimate, sports or politics, etc.)
- Categorize sentiment or behavior
- Provide predictive assessments using conditional probability
- Be trained on either the results of a query or data manually specified by users