Based on the experiments, finalize the context-aware prompting approach for LLM, including the choice of KNN algorithm, embedding strategy, vector storage solution, and prompt format.

• Scope:

  • Review findings from previous tasks (KNN, embedding, storage, etc.).
  • Finalize the approach for context-aware prompting.
  • Prepare an implementation plan for v1 based on the chosen method.

• Outcome:

  • Finalized context-aware prompting strategy for v1.
  • Implementation roadmap for the next phase.