OpenSearch Project

#OpenSearchCon 2024 North America Session Lineup is live! Come see what topics the #OpenSearch community will be speaking about, then be sure to register! 📣 Sessions: http://spr.ly/6040l9Q1p 📝 Register Now: http://spr.ly/6042l9Q1n

Explore the significant performance gains up through #OpenSearch 2.17. Explore optimizations in text search, aggregations, range queries, & vector operations, & get a preview of the OpenSearch 2025 roadmap.

Discover the power of #OpenSearch at #AWSreInvent! Visit booth 1788 to learn about scalable search, analytics, & observability. Our experts are ready to show you how to unlock insights from your data. Don't miss out!

Check out this episode of The New Stack Makers' to learn more about #OpenSearch & How the Project Went From Fork to Foundation with Anandhi Bumstead & Heather Joslyn

Explore upcoming versions of #OpenSearch & OpenSearch Dashboards using anonymous, read-only access to OpenSearch nightly playgrounds where you can try out new features & provide early feedback.

Coming to #AWSreInvent? Be sure to catch Pallavi Priyadarshini & Carl Meadows session where they will talk about #OpenSearch's journey from a fork to the creation of the OpenSearch Software Foundation under the The Linux Foundation!

Looking to improve your LLM’s response accuracy? This blog shows you how to leverage OpenSearch as a vector database for retrieval-augmented generation (RAG), giving your AI the ability to pull from a repository of domain-specific data and maintain a relevant and up-to-date LLM application. 💡 Imagine a customer service bot that references the latest internal policy or an AI assistant that offers accurate, up-to-date information every time. ➡️ Head to our blog to explore a practical example of leveraging OpenSearch in the RAG process as a retrieval tool with an experiment using a Jupyter notebook on top of Kubeflow to infer an LLM: https://lnkd.in/dbns-Zzr #OpenSource #LLM #OpenSearch OpenSearch Project Project Jupyter

벡터임베딩을 위한 Ingest Pipeline 구성하던 중 OpenSearch Project 에 첫 컨트리뷰션을 하게 되었습니다. 2.16 버전부터 InferenceProcessor의 AbstractBatchingProcessing 상속으로 배치 기능을 지원하고 있는데요, 그 동안의 OpenSearch-java에선 기능을 제공하지 않았습니다. 대부분의 경우 Data 노드가 ingest까지 맡고 있어 jvm heap 사용량을 최적화하기 위해서라도 index-level의 배치 전략을 가져가는 것이 유리할 때가 있습니다. (당장 제가 필요해서지만) java로 OpenSearch 클라이언트 개발하시는 분들께 도움이 되면 좋겠습니다. * 2.18 릴리즈에 제 commit이 포함된다고 합니다! I made my first contribution to the OpenSearch Project while configuring the Ingest Pipeline for vector embedding. Since version 2.16, batching has been supported by InferenceProcessor inheriting from AbstractBatchingProcessing, but the previous OpenSearch-java did not provide it. In most cases, the Data node is in charge of ingesting, so it is sometimes advantageous to have an index-level batching strategy to optimize jvm heap usage. I hope this will be helpful to those developing OpenSearch clients in Java. #opensearch #batch #embedding

Reminder: join us for Thursday's #OpenSearch User Group meeting in Vancouver, BC! Be sure to RSVP & we'll see you there:

Reminder: join us for Thursday's #OpenSearch User Group meeting in Vancouver, BC! Be sure to RSVP & we'll see you there: