Ray Summit 2022 Call for Papers is now open

By Jules S. Damji and Chandler Gibbons   

Ray Summit, the annual Ray community conference, is back in 2022 for a two-day, in-person event later this year (August 23-24). We are currently accepting proposals for conference talks, with a submission deadline of April 18.

In this blog post, we’ll cover the most frequently asked questions about the Call for Papers and will give you all the information you need to prepare a stellar Ray Summit proposal.

LinkWhat are the benefits of speaking at Ray Summit?

There are three main benefits to speaking at Ray Summit.

  • First, it’s an opportunity to share your Ray expertise and knowledge with the wider Ray community, giving your work wider technical recognition. This can lead to more people (including other speakers) connecting with you or seeking your advice, boosting your personal brand.

  • Second, you’ll be able to build your public speaking and presentation skills. Particularly if you plan to apply to other conferences that require previous speaking experience, speaking at Ray Summit is a great opportunity to build that experience.

  • And finally, as a speaker, you’ll get a free pass to attend Ray Summit.

LinkWhat are the key themes for this year’s Ray Summit?

The key themes and topics of interest for Ray Summit 2022 are:

  • Scalable ML: feature engineering, distributed training, or hyperparameter tuning for use cases such as graph neural networks, computer vision, time-series, and NLP/transformers

  • Reinforcement learning in the real world (recommendation systems, trading agents, etc.) built with Ray-native libraries or other frameworks

  • ML platforms or end-to-end applications: data processing through model training, serving, workflow orchestration, and monitoring

  • Deep dives into Ray core components and libraries

  • Building your own distributed framework using Ray API patterns

  • Cloud computing topics such as serverless computing and multi-cloud

  • Ray deployments on Kubernetes, on-premises, or public cloud

  • Integrations of ML libraries and frameworks with the Ray ecosystem

LinkWhat types of talks can I submit?

We have three types of talks you can submit on the above themes and topics:

  • Technical Lightning Talks (10-15 minutes): Short technical talks covering the what, why, and how of a topic with digestible code examples or quick demos.

  • Technical Talks (30 minutes): Talks that are rich in technical detail, including code examples or demos, and help the audience grasp a problem you set out to solve, challenges you met along the way, and your solution.

  • Technical Deep Dives (45-60 minutes): More in-depth, highly technical talks aimed at an advanced audience, such as an in-depth tutorial or a highly technical presentation on a deep technical topic.

For all talks, the audience should walk away having learned something new.

LinkI haven’t done this before. Any tips for preparing a proposal?

Submitting a talk to a CFP can be overwhelming, but if you focus on a few key areas, it’s easy to make sure the value of your idea shines through.

Here are a few tips for submitting a successful proposal:

  • Start with a simple and straightforward title. You only have a few seconds to grab your audience’s attention.

  • Avoid using your proposal as a sales pitch.

  • Keep it focused. You probably won’t have time to cover everything about your topic — choose a specific angle or technique to focus on.

  • Edit, edit, edit. Once you’ve written your abstract, be sure to read it over several times to make sure that it tells a clear story. Eliminate unnecessary words and sentences. Share with a trusted peer and get feedback.

  • Finally, always keep your audience in mind. Explain why people will want to attend your talk and what they’ll learn from it with three to five numbered takeaways.

For more guidance, check out the following resources:

LinkOK, I’m ready! How do I submit my talk?

Just head over to our Ray Summit page and click on Submit a talk. The submission deadline is April 18.

Next steps

Anyscale's Platform in your Cloud

Get started today with Anyscale's self-service AI/ML platform:

  • Powerful, unified platform for all your AI jobs from training to inference and fine-tuning
  • Powered by Ray. Built by the Ray creators. Ray is the high-performance technology behind many of the most sophisticated AI projects in the world (OpenAI, Uber, Netflix, Spotify)
  • AI App building and experimentation without the Infra and Ops headaches
  • Multi-cloud and on-prem hybrid support