$30 off During Our Annual Pro Sale. View Details »

Ansible Lightspeed: Where Automation Meets AI

Aykut Bulgu
September 14, 2023

Ansible Lightspeed: Where Automation Meets AI

Ansible Everywhere 2023 - Istanbul

Aykut Bulgu

September 14, 2023
Tweet

More Decks by Aykut Bulgu

Other Decks in Technology

Transcript

  1. Aykut Bulgu
    Principal Software Engineer
    @ Red Hat Ansible Engineering
    Where Automation Meets AI
    Ansible Lightspeed

    View Slide

  2. Who am I?
    #oc apply -f aykutbulgu.yaml
    apiVersion: redhat/v5.2
    kind: Principal Software Engineer
    metadata:
    name: Aykut Bulgu
    namespace: Red Hat Ansible Engineering (previously Kafka)
    annotations:
    twitter: @systemcraftsman
    email:
    - [email protected]
    - [email protected]
    organizer: Software Craftsmanship Turkey
    founder: System Craftsman
    labels:
    married: yes
    children: 2
    interests:
    - basketball/nba/rockets
    - gamification
    - stoicism
    spec:
    replicas: 1
    containers:
    - image: aykut:latest

    View Slide

  3. AI + Automation
    Choosing the right solution
    The Ansible
    Lightspeed experience
    Roadmap
    Agenda
    Intro

    View Slide

  4. Ansible is the future

    View Slide

  5. Ansible is the future

    View Slide

  6. Ansible is the future
    How close is this future?

    View Slide

  7. Ansible is the future
    Ansible Lightspeed with IBM watsonx Code Assistant
    is a generative AI service that helps developers create
    Ansible content more efficiently.
    What is Ansible Lightspeed?

    View Slide

  8. AI + Automation
    Choosing the right
    solution
    The
    Ansible Lightspeed
    experience
    Roadmap
    Ansible Lightspeed
    with IBM watsonx Code Assistant

    View Slide

  9. AI + Automation
    Red Hat wants to accelerate automation, from creation through execution
    Why? Because our customers are under increasing pressure to be
    more efficient, and turn to Ansible automation to help.
    The key to getting the most out of automation, is to drive it
    from an enterprise level. That involves multiple teams working
    together to create automation code.
    Code that scales across domains.
    Code that meets the varied needs of the teams using it.
    Code that is trusted and compliant.

    View Slide

  10. AI + Automation
    The AI revolution has come to automation
    Bridge skills gaps
    Accessible with natural language
    Expand into new automation domains
    Less time spent coding
    More focus on automations outcomes
    Better code
    Integrating generative AI into automation can greatly enhance the automation creation
    experience, helping creators produce better automation content, more quickly.
    The end result? Improved overall efficiency of an organization’s automation efforts,
    improving ROI and time to value.

    View Slide

  11. AI + Automation
    Generative AI raises a number of valid questions
    Generative AI’s potential is obvious, but in application, it comes with a
    litany of questions that need to be addressed. Among them:
    ▸ Where is generated code coming from?
    ▸ Can companies trust that their private data is protected?
    ▸ What exactly are data models being trained on - and by
    whom?
    Ansible Lightspeed with IBM watsonx Code Assistant provides
    answers to these questions - and more

    View Slide

  12. 12
    AI + Automation
    Choosing the right
    solution
    The
    Ansible Lightspeed
    experience
    Roadmap
    Ansible Lightspeed
    with IBM watsonx Code Assistant

    View Slide

  13. Choosing the right solution
    To succeed at scale with generative AI, you have to be able to trust it.
    Ansible Lightspeed with IBM watsonx Code Assistant brings
    trusted generative AI into the automation creation experience.
    It is built on three key pillars:
    Transparency Collaboration Choice

    View Slide

  14. Built to ensure transparency
    ✓ The Ansible content used to train the model is sourced from open data,
    including Ansible Galaxy, then fine-tuned with Red Hat and IBM subject
    matter expertise.
    ✓ Ansible Lightspeed will always attempt to
    match the source of each generated
    suggestion to the potential source,
    including its author and license.
    Choosing the right solution
    Collaboration
    Transparency Choice

    View Slide

  15. End-User experience
    Web Services
    Empower seasoned Ansible
    users to automate with more
    speed + efficiency
    Make it easier for automators
    of all skill levels to write
    Ansible code
    AI infrastructure + experience
    Foundation models
    Service Integrations
    Data pre + post processing
    Machine Learning
    Tools
    Model Runtime + Training
    Pipeline
    Ansible
    Lightspeed
    IBM watsonx
    Code Assistant
    Goals
    Built in collaboration
    Choosing the right solution
    Collaboration
    Transparency Choice

    View Slide

  16. Benefits for the entire DevSecOps team
    Choosing the right solution
    Collaboration
    Transparency Choice
    Automators
    Developers
    Developers get a major productivity boost
    directly in their code editing environment.
    Ansible Lightspeed is built to help
    them accelerate Ansible playbook,
    role and module creation.
    With just a basic understanding of
    YAML syntax, automators can describe
    the automation that they hope to
    create in natural language, and get
    playbook code generated in return.
    Ansible Lightspeed helps get
    automation subject matter
    expertise out of their heads - and
    into usable code.
    Operators
    Operators can have confidence that code being
    generated is compliant and reliable. Ansible
    Lightspeed is an enhancement to the automation
    platform they already know and trust.

    View Slide

  17. Ansible Lightspeed also gives users the
    opportunity to provide feedback directly
    via the VS Code extension, in order to:
    ▸ Ensure that the service is meeting user
    automation development needs
    ▸ Improve the quality and relevance of
    recommendations
    All Ansible Galaxy contributors
    have a choice as to whether
    their work is used to
    train the model.
    Choosing the right solution
    Built to deliver choice
    Collaboration
    Transparency Choice

    View Slide

  18. Choosing the right solution
    Protecting your data
    Collaboration
    Transparency Choice
    Customer data is isolated and
    will not leak into other
    customer data
    Data used to train customer models
    will only be sourced from each
    specific customer
    PII scrubbing
    Training data security
    Anonymization
    Data isolation
    Ansible best practices

    View Slide

  19. ▸ Purpose-built to accelerate the Ansible
    experience
    ▸ Uniquely positioned to offer an
    enterprise-grade AI solution for Ansible.
    ▸ Designed to deliver high-quality,
    contextual suggestions.
    ▸ Natively integrated into Ansible Visual
    Studio Code extension.
    ▸ Augments and enhances Ansible
    content creation experience.
    Non-disruptive.
    ▸ Works collectively with Ansible
    Content tools, such as Ansible Lint
    Key differentiators
    Choosing the right solution
    A singular focus
    on automation
    Meeting automators
    where they work

    View Slide

  20. Choosing the right solution
    A trusted generative AI experience, engineered for the enterprise
    Transparency
    Collaboration Choice
    Ansible Lightspeed
    with IBM watsonx Code Assistant
    A singular focus
    on automation
    Meeting automators
    where they work
    Red Hat +IBM: Ansible community | Automation SME | AI infrastructure | Enterprise expertise

    View Slide

  21. AI + Automation
    Choosing the right
    solution
    The
    Ansible Lightspeed
    experience
    Roadmap
    Ansible Lightspeed
    with IBM watsonx Code Assistant

    View Slide

  22. Capability in Tech Preview: Task Generation
    ▸ Generate tasks for a playbook or
    role from natural language task
    description
    ▸ Ansible Lightspeed suggests
    Ansible Tasks within Ansible
    Playbook and Ansible Task files
    The Ansible Lightspeed Experience

    View Slide

  23. Key feature: Content source matching
    ▸ Ansible Lightspeed will always attempt
    to clearly reference the original
    sources of content for maximum
    transparency and trust.
    ▸ Each suggestion will include potential
    source, its author and license.
    The Ansible Lightspeed Experience

    View Slide

  24. Key feature: Post-processing
    ▸ Ansible Lightspeed’s post-processing
    capability augments model suggestions
    with Ansible best practices, subject
    matter expertise, and more.
    ▸ This processing and strong contextual
    awareness stems from our unique
    insight into the Ansible code base, and
    proven experience helping customers
    automate at scale.
    The Ansible Lightspeed Experience

    View Slide

  25. AI + Automation
    Choosing the right
    solution
    The
    Ansible Lightspeed
    experience
    Roadmap
    Ansible Lightspeed
    with IBM watsonx Code Assistant

    View Slide

  26. ▸ Full playbook generation
    ▹ Generate multiple related tasks
    from a single prompt to create full
    playbooks
    ▸ Content Monitoring and
    Improvement
    ▹ Monitor Ansible content repositories
    for improvements and receive
    automatic pull requests
    ▸ REST APIs
    ▹ Integrate CI/CD pipelines,
    developer tools, and other systems
    ▸ Model Fine Tuning
    ▹ Use watsonx Code Assistant to
    create custom Ansible
    recommendation models
    Near term
    Now
    Ansible Lightspeed service offering roadmap
    ▸ Ansible Task Generation
    ▹ Use natural language in an Ansible
    Task description to generate a
    contextual suggestion.
    ▸ VS Code Extension
    ▹ Access through the Ansible VS
    Code extension
    ▸ Technical Preview
    ▹ Use the service for free through
    the technical preview
    ▹ Available now
    ▸ Content Discovery and Matching
    ▹ Find existing Ansible content instead of
    writing from scratch
    ▸ Content Description
    ▹ Generate descriptive documentation
    about the automation you build
    ▸ Content Controls
    ▹ Specific controls around data sent to
    Ansible Lightspeed
    ▸ New User Interfaces
    ▹ Web-based, CLI, and other code
    editors
    ▸ Custom Post-Processing
    ▹ Parse unstructured data into
    structures
    ▸ Runbook Recommendations
    ▹ Support recommendations for
    Event-Driven Ansible
    Long term
    Roadmap

    View Slide

  27. Ansible Lightspeed
    Ansible Lightspeed with IBM
    watsonx Code Assistant:
    Technical Preview
    Ansible Lightspeed with IBM
    watsonx Code Assistant:
    Commercial version
    Access VSCode extension
    VSCode extension
    REST APIs
    Git Repository Scanning
    Capabilities Task generation
    Content generation
    Content discovery
    Content optimization
    Content explanation
    Data sets Ansible Galaxy
    Watsonx.ai Foundation Models +
    custom data model import
    Vendor Red Hat Red Hat & IBM
    Pricing Free Per seat
    Availability June 2023 Q4 2023

    View Slide

  28. Next Steps
    redhat.com/ansible-lightspeed
    ▸ Get started with the service
    ▸ Community docs
    ▸ Technical Preview video
    ▸ Technical preview blog

    View Slide

  29. linkedin.com/company/red-hat
    youtube.com/user/RedHatVideos
    facebook.com/redhatinc
    twitter.com/RedHat
    Thank you

    View Slide