Introduce RAG for personalization and history

[MrCsabaToth]

This needs to work together with function calls and ReAct #5 / ReWOO #6

[MrCsabaToth]

Since objectbox (https://pub.dev/packages/objectbox) supports vector DB as well, it'd cover both history (#8) and RAG

[MrCsabaToth]

Note, once the basic RAG is in place, explore:

• different embeddings
• hybrid-search (applicable?)
• auto-cut (crop documents when the score suddenly drops)
• re-rank (applicable?)

[MrCsabaToth]

https://ai.google.dev/gemini-api/docs/embeddings#dart-flutter

final model = GenerativeModel(model: 'text-embedding-004', apiKey: apiKey);
final content = Content.text('The quick brown fox jumps over the lazy dog.');
final result = await model.embedContent(content);
print(result.embedding.values);

dimensionality is 768

[MrCsabaToth]

We introduced thresholds as well now.