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

Introducing Anyscale: Accelerate scaling and deployment of AI / ML

Anyscale
PRO
December 16, 2021

Introducing Anyscale: Accelerate scaling and deployment of AI / ML

We are excited to announce the general availability (GA) of the Anyscale managed Ray platform.

We created Ray to enable every developer to effortlessly scale their AI/ML workloads from the laptop to the cloud without needing to built and manage complex infrastructure, or become a distributed systems expert. The Anyscale platform builds on that vision and brings the power of instant scaling on Ray to a much larger community.

Join our special Anyscale GA launch event with Ray creators and Anyscale founders, Robert Nishihara and Ion Stoica, for an overview and live demo.

Session highlights include:
- Infinite laptop experience: Accelerate development and experimentation through instant access to cloud scale with a laptop-like experience.
- Production jobs & services: Effortlessly deploy AI workflows and models into production
- Unified compute: Manage the full AI lifecycle on a single platform

Anyscale
PRO

December 16, 2021
Tweet

More Decks by Anyscale

Other Decks in Technology

Transcript

  1. Introducing Anyscale:
    Accelerate the scaling and
    deployment of AI applications
    Ion Stoica | Anyscale
    Bill Chambers | Anyscale
    Tobias Kaymak | Ricardo

    View Slide

  2. Exciting News!
    Anyscale is now
    generally available
    $100M Series C
    funding round

    View Slide

  3. 87%
    Market Problem
    Source: Venturebeat 2019
    AI projects don’t make it
    from pilot to production

    View Slide

  4. ANYSCALE MISSION
    Accelerate the scaling and
    productionization of AI apps

    View Slide

  5. Ray Mission
    We created Ray to make it effortless to scale any
    workload from your laptop to the cloud.

    View Slide

  6. Amazing growth in 5 years
    600+
    18000+
    1000s
    Contributors
    GitHub stars
    organizations are using Ray
    25+
    Libraries

    View Slide

  7. Introducing Anyscale
    The best way to scale and deploy AI applications on Ray

    View Slide

  8. Accelerate the development and deployment of
    AI applications at any scale
    Develop on a laptop
    Instantly scale to the cloud

    View Slide

  9. Edit
    Run
    Debug
    Test /
    staging
    Deploy
    Development Production
    AI Application Lifecycle

    View Slide

  10. Edit
    Run
    Debug
    Test /
    staging
    Deploy
    Development Production
    AI Application Lifecycle

    View Slide

  11. `
    Run
    Debug
    Test /
    staging
    Deploy
    Development Production
    Anyscale: Best of both worlds
    Laptop development experience and cloud scale
    Edit

    View Slide

  12. INFINITE LAPTOP EXPERIENCE
    Accelerate AI development
    with instant access to cloud
    scale from your laptop
    Write once. Run anywhere.
    Scale from laptop to cloud with zero code
    changes
    Simple Python APIs
    Compose distributed applications in
    native Python code
    Distributed ML libraries ecosystem
    Unify the end-to-end AI lifecycle on a
    single, scalable platform
    Dependency packaging
    Auto sync local environment and files to
    the cloud

    View Slide

  13. INFINITE LAPTOP EXPERIENCE
    Accelerate AI development
    with instant access to cloud
    scale from your laptop
    Write once. Run anywhere.
    Scale from laptop to cloud with zero code
    changes
    Simple Python APIs
    Compose distributed applications in
    native Python code
    Distributed ML libraries ecosystem
    Unify the end-to-end AI lifecycle on a
    single, scalable platform
    Dependency packaging
    Auto sync local environment and files to
    the cloud
    Edit
    Run
    Debug
    Develop, at scale, on a laptop

    View Slide

  14. PRODUCTION-READY
    Effortlessly deploy AI models
    & workflows into production
    Production jobs & services
    Deploy ML workflows & models into
    production with ease
    Serverless experience
    Elastic scaling with no runtime limits or
    GPUs restrictions
    CLIs, APIs & SDKs
    Automate and integrate into your
    workflows (eg. CI/CD)
    Observability & alerting
    Track health with event logs and prebuilt
    dashboards
    Run
    Test /
    stage
    Deploy
    Edit
    Debug

    View Slide

  15. PRODUCTION-READY
    Effortlessly deploy AI models
    & workflows into production
    Production jobs & services
    Deploy ML workflows & models into
    production with ease
    Serverless experience
    Elastic scaling with no runtime limits or
    GPUs restrictions
    CLIs, APIs & SDKs
    Automate and integrate into your
    workflows (eg. CI/CD)
    Observability & alerting
    Track health with event logs and prebuilt
    dashboards

    View Slide

  16. ANYSCALE ADVANTAGE
    Amplify your Ray journey on a
    managed platform from the
    creators of Ray
    100% managed infrastructure
    Eliminate your operations load with a
    fully-managed service
    Multi-cloud enabled
    Run the same code on any cloud with no
    code changes.

    View Slide

  17. DEMO

    View Slide

  18. Scaling AI with Anyscale
    Conversation with Tobias Kaymak,
    Sr. Data Engineer, Ricardo

    View Slide

  19. Ray Meetup
    Jan 19 - Virtual / SF
    Join the Ray Community
    ray.io/community
    Slack Discuss Forum
    discuss.ray.io

    View Slide

  20. SIGN UP TO REQUEST ACCESS
    anyscale.com/signup
    Reduce the time to market on your AI projects
    with Anyscale and Ray

    View Slide

  21. Thank you!

    View Slide