About
AI agents often suffer from a lack of persistent memory, leading to repeated mistakes and loss of context between sessions. Shodh-Memory addresses this by offering a robust cognitive memory system that learns from experience and stores information across sessions. It implements a 3-tier memory architecture inspired by Cowan's model, featuring Hebbian learning for strengthening connections, activation decay for forgetting, and semantic consolidation. Packaged as a lightweight, single binary, Shodh-Memory runs entirely offline, making it ideal for cloud, edge devices, or air-gapped systems, ensuring privacy and sub-millisecond retrieval speeds.
