[Feature] Support Elasticsearch as memory vectorDB

[dej611]

Description

This is an attempt to add an adapter for the memory storage in order to let the user configure a different vectorDB than the bundled Qdrant one.
I have 0 experience with python, so sorry about silly mistakes in this PR.

My idea was to start with something simple like configure it via .env file and the transition for the final user should be completely transparent.
I guess a setting UI would be a nicer experience here, keeping the API_KEY approach.

Features planned:

• ability to configure an ES host
  • Test it with a local instance (HOST + PORT + API_KEY)
• support cloud instance
  • Test it with a cloud instance (CLOUD_ID + API_KEY)
• abstract the VectorMemory logic with an adapter
  • remap all the collection logic into ES-specific logic
  • provide lean adapters for ES results
  • provide basic knn search
  • handle the switch embedder scenario
  • implement dump logic

Related to issue #1039

Type of change

• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas