Upgrade to Pro — share decks privately, control downloads, hide ads and more …

How AI Powers DevOps and ITSM

How AI Powers DevOps and ITSM

Helen Beal

August 28, 2024
Tweet

More Decks by Helen Beal

Other Decks in Technology

Transcript

  1. Internal 2 Helen Beal Helen Beal is Head of the

    Ambassador Program at PeopleCert (for DevOps Institute, ITIL, PRINCE2 and LanguageCert). She’s the CEO and chair of the Value Stream Management Consortium and co-chair of the OASIS Value Stream Management Interoperability Technical Committee, and chair of the DevNetwork DevOps Advisory Board. She also provides strategic advisory services to DevOps and VSM industry leaders. She is the author of the annual State of VSM Reports from the VSMC and the State of Availability Reports from Moogsoft. She is a co-author of the book about DevOps and governance, Investments Unlimited, published by IT Revolution. She is a DevOps editor for InfoQ, and also writes for a number of other online platforms. Helen hosts the Day-to-Day DevOps webinar series for BrightTalk and serves on advisory and judging boards for many initiatives including Developer Week, DevOps World, JAX DevOps, and InterOp. Bringing joy to work
  2. Internal What’s the relationship? 3 DevOps ITSM Focus The rapid

    and reliable delivery of software and services with an emphasis on collaboration, automation, and continuous improvement throughout the entire software development lifecycle. The effective management and delivery of IT services to meet business needs with an emphasis on stability, control, and ensuring IT services align with organizational goals. Methodology An agile and iterative approach, with frequent releases and a focus on continuous feedback and improvement. It encourages collaboration between development, operations, and other teams to streamline the software delivery process. Utilizes a more structured and process-oriented approach, often following frameworks like ITIL. It places emphasis on defined processes, service level agreements (SLAs), and incident management to ensure stability and reliability. Goals Accelerate software delivery, improve quality, and increase efficiency. It prioritizes faster time to value, continuous innovation, and customer satisfaction. Ensure IT services are aligned with business needs, deliver value, and maintain high levels of availability and performance. It prioritizes stability, cost-effectiveness, and customer satisfaction. Culture Collaboration, shared responsibility, and continuous learning, encouraging breaking down silos between teams and fostering a sense of ownership throughout the software development lifecycle. Process adherence, control, and accountability with an emphasis on defined roles and responsibilities, clear communication, and service-level management. Tools Utilizes a variety of tools to automate and streamline the software delivery pipeline, such as CI/CD tools, configuration management tools, and containerization platforms. Leverages tools for service desk management, incident management, change management, and asset management to track, control, and optimize IT services.
  3. Internal 5 Types of AI Dimension Creation Learning Approach Output

    Adaptability Applications Traditional Primarily focuses on analyzing existing data to classify, predict, or make decisions based on patterns within that data. Typically relies on supervised learning, requiring labeled data to train algorithms for specific tasks. Provides predetermined outputs based on specific rules and patterns learned from training data. May struggle to handle scenarios outside its predefined training data, requiring retraining for new tasks or situations. Primarily used for tasks like classification, prediction, optimization, and automation in various industries. Generative Its primary goal is to create new, original content – text, images, music, even videos – that often mimic human creativity. Leverages unsupervised or semi-supervised learning techniques to discover patterns and structures in data without explicit instructions. Generates novel and diverse outputs, often exhibiting a degree of creativity or artistry. Can adapt to new situations and generate content even when presented with unfamiliar prompts or data, exhibiting a degree of generalization. Used in creative fields like art, music, and writing, as well as for content creation, design, and even drug discovery.
  4. Internal 12 If your organization is investing or planning to

    invest in AI initiatives and tools, where did the requirement originate?
  5. Internal 16 The Digital Delivery Lifecycle CONTINUOUS INTEGRATION Code is

    created, artifacts incorporated, versions controlled, code is built in a trunk based manner. CONTINUOUS DELIVERY The changes are approved, released and operated in the live environment. CONTINUOUS TESTING Functional and non-functional testing takes place at every commit at every step or gate through route to live. PORTFOLIO AND BACKLOG Vision and goals are set and aligned to epics, features, PBIs and user stories. INSIGHTS AND ANALYSIS Monitoring and observability provide insights into customer reaction to changes and report on value realization. DevOps Toolchain
  6. Internal 17 The Digital Delivery Lifecycle - AI Powered CONTINUOUS

    INTEGRATION Code generation, completion, refactoring and optimization, intelligent code search and navigation, NLP for code understanding, personalized development assistance, build optimization CONTINUOUS DELIVERY Predictive success and failure analysis, intelligent release orchestration, dynamic release decision making, autonomic remediation CONTINUOUS TESTING Test case generation and prioritization, smart test execution and result analysis, bug detection, static and dynamic code analysis PORTFOLIO AND BACKLOG Intelligent prioritization, predictive analytics and risk management, resource optimization, automated work item classification and assignment, performance tracking and reporting INSIGHTS AND ANALYSIS AIOps, self-service chatbots, NLP powered sentiment analysis, personalization and recommendation engines, customer journey mapping AI Use Cases
  7. Internal 18 Portfolio and Backlog Intelligent Prioritization Predictive Analytics Resource

    Optimization Automated Task Management Data-Driven Decision Making AI algorithms can analyze vast amounts of data related to projects, including dependencies, risks, resource constraints, and strategic alignment. This allows for automated prioritization of backlog items based on their potential value, urgency, and strategic importance. AI models can predict project outcomes, such as completion dates, resource utilization, and potential risks. These insights enable proactive management, risk mitigation, and better resource allocation. AI can help identify the optimal allocation of resources across multiple projects. This includes matching skills to tasks, avoiding overallocation, and anticipating resource bottlenecks. AI-powered tools can automate routine tasks, such as progress tracking, status updates, and data entry. ChatGPT can automate repetitive tasks such as drafting emails, creating project summaries, and generating status reports. This frees up time to focus on more strategic activities. AI algorithms can provide actionable insights based on historical project data and real-time performance metrics. This enables informed decision-making, portfolio balancing, and strategic adjustments.
  8. Internal Code Generation and Completion Refactoring and Optimization Intelligent Code

    Search NLP for Code Understanding Intelligent Build Management Tools like Copilot suggest code snippets, complete functions, and even generate entire classes based on the context of the code being written. AI models can generate code from natural language descriptions, allowing developers to express their intent in plain language and have the AI translate it into functional code. AI-powered tools can analyze codebases to identify areas where refactoring can improve code quality, maintainability, and performance. They can pinpoint code smells, complex structures, redundant code, and potential performance bottlenecks. They can suggest refactoring actions and even automate them. Traditional code search often relies on keyword matching, which can lead to irrelevant results. AI-powered code search, on the other hand, leverages natural language processing (NLP) and machine learning to understand the context and intent behind a search query. This allows it to retrieve more relevant code snippets, even if they don't contain the exact keywords used in the search. AI can analyze code and generate documentation automatically, reducing the manual effort required to keep documentation up to date. AI-powered tools can facilitate communication among team members, provide contextual information during code reviews, and even translate technical jargon into plain language. AI can optimize build configurations and dependencies, reducing build times and improving efficiency and it can analyze build failures and provide insights into the root causes, helping developers resolve issues faster. 19 Continuous Integration
  9. Internal Test Case Generation Smart Test Execution Bug Detection Code

    Analysis Security Testing AI can analyze code changes, requirements, and user stories to automatically generate relevant test cases, increasing test coverage and reducing the manual effort involved. AI can prioritize test cases based on their criticality, risk, and potential impact, ensuring that the most important tests are executed first and resources are allocated efficiently.AI-powered test automation tools can adapt to changes in the application's UI or behavior, reducing the fragility of test scripts and minimizing maintenance efforts. AI analyzes test failures, identifies root causes, and even suggests fixes, accelerating debugging and remediation. AI-powered static code analysis tools can examine the code without executing it, looking for patterns that indicate potential bugs, security vulnerabilities, or performance issues. AI can also be used to analyze code as it is running, looking for anomalies that might indicate a bug. AI can analyze vast amounts of security data (logs, network traffic, user behavior) to detect anomalies and suspicious activities that might indicate a security breach or attack for proactive threat identification. It can analyze threat intelligence feeds from various sources to stay up-to-date on the latest vulnerabilities and attack vectors. 20 Continuous Testing
  10. Internal Predictive Success & Failure Analysis Intelligent Release Orchestration Dynamic

    Release Decision Making Autonomic Remediation Intelligent Pipeline Optimization* AI can analyze historical deployment data, application performance metrics, and other relevant factors to predict the success or failure of a deployment. This enables teams to make informed decisions about when and how to release new versions of software. AI can analyze real-time data from a small subset of users (canary group) to assess the impact of new features or changes before rolling them out to the entire user base. This allows for early detection of issues and controlled release management. AI can continuously monitor and analyze a vast array of real-time data sources related to software releases and utilize this data to predict potential issues and the impact of a release on customer experience, helping organizations plan their release calendar. In case of a failed deployment, AI can automatically trigger a rollback to a previous stable version of the software, minimizing downtime and impact on users. AI can analyze pipeline execution data to identify bottlenecks, inefficiencies, and potential areas for improvement. It can then suggest optimizations or even automate the process of improving the pipeline's performance and reliability. 21 Continuous Delivery *Digital value stream management
  11. Internal 23 Insights and Analysis AIOps Self-Service Chatbots NLP Powered

    Sentiment Analysis Personalization & Recommendations Customer Journey Mapping By automating incident detection and providing actionable insights, AIOps enables IT teams to respond to and resolve issues more quickly, minimizing downtime and impact on users. AIOps can predict and prevent potential problems before they cause service disruptions, ensuring higher system availability and performance. AIOps can predict and prevent potential problems before they cause service disruptions, ensuring higher system availability and performance. By providing self-service options and guiding users to relevant knowledge articles or FAQs, chatbots can deflect a significant number of tickets from reaching the IT service desk, reducing agent workload. AI powers NLP sentiment analysis by leveraging machine learning algorithms and natural language processing techniques to interpret and categorize human emotions and opinions expressed in text. MLmodels are trained on labeled data to learn the patterns and associations between words, phrases, and sentiment categories. AI profiles users and segments them. Using historical data and ML algorithms, it predicts user needs and preferences. It applies sentiment analysis and context (location, device, time of day) to provide relevant, timely recommendations - in ITSM for service requests and incident resolutions. For customer experience, additional products and services or customer support. AI can gather and analyze data from a wide range of customer touchpoints, including websites, social media, mobile apps, customer support interactions, and even physical store visits. This creates a comprehensive view of the customer journey, allowing businesses to identify patterns and trends across multiple channels and the omnichannel.
  12. Internal 24 Value Stream Management PERFORMANCE ANALYSIS Pattern detection in

    flow metrics, identify optimization and improvement opportunities WORK ITEM PRIORITIZATION Ranking work based on customer and business value and outcomes BOTTLENECK DETECTION Predictive analytics and ML algorithms to identify waste, risks, delays and dependencies MAPPING AND VISUALIZATION Builds map based on data from DevOps toolchain including ITSM tools CONTINUOUS IMPROVEMENT Measures progress, automates repeated tasks, relieves cognitive load, injects innovation capacity A digital value stream
  13. Internal 27 GenAIOps Lifecycle Management DataOps LLMOps DevOps ModelOps Security

    Management Governance, Risk and Compliance Ethics AI powers the delivery of AI solutions.
  14. 28