It was a lot of fun building this initial RAG implementation with the @atolio team! Of course we weren't starting from scratch like most folks. Our ingestion pipeline, integrated permissions, and @vespaengine core made the retrieval side smooth. A good search core == data ops

R to @dainius_jocas: Maybe loading from zip files instead of plain text files would be most useful. Let's see

R to @jobergum: So far using the path works like magic! What Uwe proposes is an extension to where the resource files can be loaded from, like S3. I'm also not entirely sure if that would offer something more than what it is possible to achieve now, but that would be a new thing to learn for me.

R to @thetaph1: The Vespa path configuration allows developers to ship files in the application package and use these at runtime in custom components. The path is relative to the application package. Vespa distributes the files to all the serving containers/JVMs.

R to @thetaph1: Thanks! I’ll try to experiment how a custom ResourceLoader could plug into the machinery.

RT by @johtani: Vespa (@vespaengine) で利用できる Lucene ベースの Linguistics plugin の話。

R to @jobergum: Cool ideas with using CustomAnalyzer! Instead of using Path, you could have implemented your own ResourceLoader. Then data files could also be loaded from other places like zip files, sine internal database blob store,...

RT by @michaelcizmar: This is big for those migrating to @vespaengine from engines based on Apache Lucene (Elasticsearch/OpenSearch/Solr). From @dainius_jocas https://blog.vespa.ai/lucene-linguistics/

R to @jobergum: Thank you for the initial inspiration!

R to @vespaengine: vespa does seem like an extremely valid choice compared to @qdrant_engine

R to @kraune: 3/ Thanks to @dainius_jocas and his team for creating this, and try out some the examples added to the @vespaengine sample applications library at https://github.com/vespa-engine/sample-apps/tree/master/examples/lucene-linguistics

Finally, a @vespaengine blog post introducing #lucene linguistics is live! 🛵 Try it out and let me know how it goes. https://blog.vespa.ai/lucene-linguistics/

I’m biased but @vespaengine, and we are hiring developer advocates https://vespa.ai/we-are-hiring

Feeling great! 😎 My little experiment has contributed to the @vespaengine documentation!

R to @o19s: We also had @jobergum from the @vespaengine team on "Navigating Neural Search: Avoiding Common Pitfalls" https://youtu.be/PrBtxqARY9U [3/5]