Hey everyone,
Was wondering if anyone has successfully implemented long-term memory with intelligent updating for AI agents?
Similar to how ChatGPT remembers context and updates based on new inputs.
Specifically, I’m looking for:
- A way to store persistent memory that evolves over time.
- Intelligent updating of existing data—e.g., overwriting outdated information without losing relevant context. (AI agent specifically for this?)
Has anyone cracked the code on this? Would love to hear how you’ve approached it, what the workflow looks and any pitfalls to avoid.
Appreciate any insights or examples you’re willing to share!