This example demonstrates how a Python AI agent can use a simulated vector database as its direct, continuously updated memory, rather than just for static information retrieval. It shows the agent learning from new experiences, recalling relevant past interactions, and using this memory to inform its responses and learn from the interaction itself.
python
Save the code as main.py. Run from your terminal: python main.py.
This example accompanies the Turkish article: Python Ajanları İçin Vektör Veritabanını Hafıza Olarak Kullanmak: Geri Çağırma (Retrieval) Mekanizmasının Ötesinde Bir Yaklaşım.
MIT — see LICENSE.