{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/792ce1a6b73c4fc0bd0c5696a8bd6b60\" frameborder=\"0\" width=\"1316\" height=\"987\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":987,"width":1316,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":987,"thumbnail_width":1316,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/792ce1a6b73c4fc0bd0c5696a8bd6b60-f214bac154b7701d.gif","duration":746.628,"title":"Exploring Memory Strategies in Conversational AI 🤖","description":"In this video, I present my memory comparison project, where I explore three different memory strategies for a Langchain conversational agent: baseline, session memory, and long-term memory. The baseline strategy stores no information, while session memory retains context within a single thread, and long-term memory allows for cross-thread recall of key facts. I demonstrate these strategies through three scripted conversations, highlighting their distinct behaviors and storage mechanisms. I encourage you to consider how these memory strategies could be applied in real-world scenarios, such as customer support or personal AI assistants. Please take a look at the code and tests I implemented to see how these strategies function in practice."}