Weaviate Newsletter - Weaviate 1.15, WCS and new content

September 14, 2022

Dear Weaviate friend,

We had an exciting couple of weeks since the last newsletter.

We released a new version of Weaviate v1.15, announced a new pricing model for the Weaviate Cloud Service (WCS), published new pieces of content (blogs and podcasts), and opened up new job opportunities.

Weaviate v1.15

🎉 The list of new features and improvements that went into Weaviate 1.15 was so long, that the TL;DR could have its own TL'DR. 😉

✍️ You can read all about it in the release post, or jump straight to a specific section from the below list:

🎙 If you prefer your content in a spoken-word format, then tune in to the release podcast with Connor, Etienne, and Drik - and let them explain to you what is new in Weaviate and why all of that is very important.

What do you think of the latest release? Please feel free to share your thoughts and feed to this email or join the conversation on Slack.

Weaviate Cloud Service (WCS)

If you have never heard of the Weaviate Cloud Service, then I have a great surprise for you. WCS allows you to spin up an instance of Weaviate in the cloud and communicate with it just like you do with locally deployed docker instances.

You can test it out with a free sandbox cluster. Head over to the Weaviate Console, create a WCS account, and create a sandbox instance. From now on, you can communicate with your cluster via any Weaviate language client or with the REST API.

Note: Although the sandbox configuration doesn't let you pick your modules, it comes packaged up with Hugging Face and OpenAI modules. So, you can pick any of your favourite sentence transformers and query away 😉

The announcement

Inspired by “pay-as-you-grow” pricing used in cloud storage, we introduced a new business model that makes it easier and more affordable for enterprises of any scale to take advantage of its Weaviate vector-search engine. Users have the option of paying a usage-based rate for search on a per-dimension basis. Pricing begins at $0.05 per million dimensions.

✍️ You can find the updated pricing on our website and read the press release from Bob (CTO at SeMI).

Content

✍️ We often get asked, "why is Weaviate so fast?". You can find the answer in Why is Vector Search so fast - by our own Laura.

✍️ Parker (one of our core engineers) has been around for over 6 months. He shared his excitement and experience in using Real World Algorithms on LinkedIn.

✍️ Our in-house designer Svitlana, wrote about Building an open-source brand - where she shares how we do it at SeMI Technologies. It is always fascinating to see the world from a designer's point of view.

✍️ Read The New World of Vectorization and AI-First - Sam Ramji, to learn how we can build the future of intelligent data through an open-source AI ecosystem.

✍️ If you are interested in the inner works of HNSW, then IVFPQ + HNSW for Billion-scale Similarity Search by Peggy Chang is a must-read.

❓We are always looking for new ideas to write and talk about. Are there any topics that you are interested in that you would like us to cover? Or would you like to write something yourself? Give me a shout, I can help.

Who to follow

If you are a "blue bird" enthusiast, who likes to consume your content via Twitter, then you should follow Erika Cardenas (@ecardenas300). Every Saturday morning, Erika shares a summary of all that happened around Weaviate that week.

Updates

We have a few other updates that go with Weavaite v1.15.

Language clients

Weaviate Helm v15.0.0 – release notes

This version adds support for:

  • the new backup modules: backup-filesystem, backup-s3 and backup-gcs

  • the new modules: sum-transformers and text2vec-huggingface

Jobs

We would like to share the Community Solution Engineer vacancy. This is an exciting role for anyone that likes to work with the community and loves building real-life applications. If that is you, apply here.

We have other job opportunities as well. Check them out here, who knows maybe we could work together 🤗

Thank you for reading,

Sebastian