feat: evaluate vector storage: in-memory vs vector DB

[iamanishsrivastava]

Compare in-memory vector storage with vector DB options to assess which one provides better performance for context retrieval during LLM interactions.

• Scope:

  • Set up in-memory vector storage and vector DB.
  • Test both methods for speed, scalability, and retrieval accuracy.
  • Provide a performance comparison with documented recommendations.

• Outcome:

  • Performance data comparing in-memory storage vs vector DB.
  • Clear recommendation for the vector storage approach.

[github-actions]

This GitHub issue proposes assessing in-memory vector storage against vector databases to determine the optimal solution for context retrieval in LLM interactions. The scope involves setting up both storage methods, evaluating them on metrics like speed, scalability, and retrieval accuracy, and documenting the findings. The intended outcome is to produce performance data comparing the two approaches and provide a clear recommendation on which storage method is better suited for the use case.