AgentKeeper – cognitive persistence layer for AI agents
Cross-provider agent memory is clever, but LLM context windows keep growing and RAG is already standard.

Knowledge graph compression (3,714x token ratio) is impressive, but 'persistent agent memory' is crowded territory.
Enterprise AI teams shipping production agents, especially healthcare/finance
Pinecone · Weaviate · LlamaIndex
It's a cognitive memory layer that gives agents persistent recall, automatic fact extraction, and temporal reasoning — via a single SDK call.
Key numbers: - 89% accuracy on LoCoMo long-term conversation benchmark - 50ms p50 retrieval latency - 97% cost reduction vs raw token re-processing
Open source SDKs, works with any LLM stack. Currently in early access.
Happy to answer questions about the architecture, benchmark methodology, or how we handle knowledge graphs.
Cross-provider agent memory is clever, but LLM context windows keep growing and RAG is already standard.
Recovers 95% critical facts when switching GPT-4 ↔ Claude with real benchmarks.
Claims brain-like cognition with zero LLM calls, but zero evidence of actual learning.
SQLite-backed agent memory with graph viz when Mem0 and Zep already dominate.
Bundled 7MB embedder means zero network calls or model downloads for agent memory.
Trigger-based cognitive architecture for Claude Code loses context anyway without API-level state persistence.