This week in Jina AI

Announcing Jina 3.2

Weaviate, Qdrant, GraphQL, switches in Flows, and more…

All New Features in Jina 3.2

Our latest update is packed with features from DocArray and improves its readiness to be run in docker-compose.

• Now you can add conditions into your Flow to build very complex data pipelines, leveraging Query Language support provided by DocArray.
• Health checks: enable health checks for containers when dumped into docker-compose files, including gateway using http and websockets protocols.
• Now you can set the type of ndarray output for Executors for increased interoperability of Executors inside a Flow.
• Added GraphQL support: we support GraphQL queries in our HTTP gateway which will allow you to define the returning schema of Documents. You can now select fields of the Documents that are most relevant to your application.
• on_error callback now gets access to exceptions raised from the Client.
• UI/UX: We’ve unified port_expose and port_in into a single parameter port to make your life easier. This change is backward compatible.
• UI/UX: Added isort and flake8 for automatic import sorting
• UI/UX: Documentation and UI improvements to better clarify the usage of client, Flow API, etc. But don’t just take our word for it, try it here!

All New Integrations

Weaviate

We’re integrated DocArray with Weaviate!

You can now leverage our robust Document Store with the same DocumentArray API, Document retrieval on embeddings in the blink of an eye, and lower latencies at any scale!

Watch it in action here!

QDrant

Your vector search applications in DocArray can now be backed by Qdrant. Add one line of code to your DocumentArray and enjoy the lightning-fast vector retrieval delivered by a local/remote Qdrant backend. Learn more✨

DocArray 0.9.14

Our latest version of DocArray is packed with some super cool features.

• You can now call a local/remote Flow or Hub Executor/Sandbox like a regular function. Try it out now on the Jina Hub
• Using find(), you can now select Documents from a DocumentArray based on the conditions specified in a query object. You can even use da.find(query) to filter Documents and get nearest neighbors from da. Learn more about how this can be achieved.
• By using post(), you can process a DocumentArray by calling an external Flow/Sandbox/Executor, which may be either local, remote, or inside a Docker container. Learn more about the post() feature.

Sandbox

With Jina’s Sandbox, you can evaluate Executors faster to find the right one for you. Just use jinahub+sandbox://ExecutorName, and you can run that Executor on Jina’s cloud without using local compute.

• We’ve added a 3.0 Exclusive tag to help you pick the Executors supported with our latest version.
• You can now see the number of sandboxes available next to an Executor and the number of downloads.

We’ve built a demo for you to experience the full power of Jina Sandbox. Check it out here!

Upcoming Events

Neural Search with Jina AI [online workshop]

Join our Product Manager, Susana Guzman, for an introductory workshop on neural search with Jina AI on Monday, March 28th at 19:00 CET (👉 Convert to your time zone) with the brilliant team of PyLadies Berlin.

Sign up now!

Office Hours

We’re hosting our second installment of office hours on Thursday, March 31st at 17:00 CET (👉 Convert to your time zone). If you have questions, want to discuss projects, or just want to say hi, feel free to drop by!

Sign up now!

Latest Collaborations

Weaviate Podcast #11: CEO Han Xiao from Jina AI

Weaviate Integration Tutorials

Women’s History Month

Get Involved!

Contribute to our Good First Issues Repo

We have launched our good first issues repo for those looking to take their first steps into open source with Jina. Check out these issues:

• Example: NFT/blockchain search engine
• Update Jina Hub Executors for Jina 3.0+
• React component for Jina

Get Jina Certified

Take our Learning Bootcamp quiz with beginner, intermediate and advanced tracks to etch your name in our Hall of Fame!

Join our Slack Community!

Our community now boasts over 2,700 neural search enthusiasts. We would love to invite you to join us!Join us!✨

Join our Meetup Group

Join our meetup group to stay updated with our latest events.

Whether you’re AI engineers, Python Engineers, Data Scientists, CTO, Machine Learning Researchers, Open Source Evangelists, AI Product Managers, or just plain old interested, we’d like to meet up with you!