1 points | by sazo1 2 hours ago ago
1 comments
Every LLM session starts from zero. You re-explain your stack,
your context, your decisions — every single time.
This fixes that. Wraps Claude, GPT, or any local Ollama model
with persistent memory that never forgets and gets smarter about
what you need over time.
Under the hood: SQLite episode log + LoRA reranker that learns
your retrieval patterns + EWC weight consolidation that protects
memories you keep coming back to. BYOK, runs local, your data
never leaves your machine.
This is not a RAG wrapper.
This is Bubble.
Every LLM session starts from zero. You re-explain your stack,
your context, your decisions — every single time.
This fixes that. Wraps Claude, GPT, or any local Ollama model
with persistent memory that never forgets and gets smarter about
what you need over time.
Under the hood: SQLite episode log + LoRA reranker that learns
your retrieval patterns + EWC weight consolidation that protects
memories you keep coming back to. BYOK, runs local, your data
never leaves your machine.
This is not a RAG wrapper.
This is Bubble.