AI Agent Memory System Integration Solution Evaluation and Selection Prompt

You are an expert AI systems architect specializing in agent memory infrastructure. I need you to help me evaluate and select the right memory system for my AI agent project. ## My Requirements - Agent framework: [e.g., LangChain / CrewAI / custom Python agent] - LLM provider: [e.g., OpenAI / Anthropic / local models] - Scale: [e.g., single agent / multi-agent team / production SaaS] - Memory needs: [e.g., conversation history / learned facts / cross-session persistence / cross-agent sharing] - Deployment: [e.g., cloud API / self-hosted / hybrid] ## Please Analyze 1. **Feature Matrix**: Compare Memori, mem0, memsearch, and Letta (MemGPT) across: persistence model, cross-agent memory sharing, structured vs unstructured recall, LLM-agnostic support, SDK languages, and pricing. 2. **Architecture Fit**: Based on my requirements, recommend the top 2 options with reasoning. 3. **Integration Effort**: For each recommendation, provide a step-by-step integration plan with estimated effort (hours). 4. **Migration Path**: If I start with one and need to switch later, what is the data portability story? 5. **Production Checklist**: Key considerations for taking agent memory to production (latency, cost, privacy, backup). Output as a structured report with comparison tables and actionable next steps.