Andy Flury

thx @langchain4j! I did end up using the same memory for multiple AI services, which seems to work. In case anybody else needs this here is my code: ``` @Service...

Thank you @langchain4j Yes, that's exactly the use case we have in mind. And yes, adding a tool (or tools) to assistant calls would be ideal. Is this already possible...

Yes, the tools would be selected by users. We are thinking about adding a tool selection to our chat UI. At the moment we have about half a dozen tools,...

I think the concept with the `EmbeddingModelTextClassifier` is great. Main problem with letting LLM's decide about Tools is that you have to send tool definitions with every LLM call which...

Interesting! @langchain4j So this LLM based approach would result in two LLM calls for each user request? The first containing the list of tools with short description and the second...

Would also very much appreciate to have meta data filtering. Spring AI has something like this already, the call it [metadata filters](https://docs.spring.io/spring-ai/reference/api/vectordbs.html#_metadata_filters) Our use case is that we index different...

We are using ElasticSearch on our side. Thx!

As I quick workaround I created my own extension of `ElasticsearchEmbeddingStoreWithFilter`, which overwrites `buildDefaultScriptScoreQuery` as follows ``` private ScriptScoreQuery buildDefaultScriptScoreQuery(float[] vector, Query query, float minScore) throws JsonProcessingException { JsonData queryVector...

I'm also not an Elasticsearch expert. I basically copied this logic here: https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-script-score-query.html#vector-functions-cosine

This looks great @langchain4j Thanks for the quick turnaround! I just tested in with my application with a simple `equals` comparison and it worked like a charm! Regarding your questions....