<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>PLUR Blog</title><description>Open engram format, agent memory architecture, and PLUR Enterprise.</description><link>https://plur.ai/</link><language>en-us</language><item><title>Mem0 vs Letta vs Zep: Which Should You Use for Agent Memory?</title><link>https://plur.ai/blog/mem0-vs-letta-vs-zep/</link><guid isPermaLink="true">https://plur.ai/blog/mem0-vs-letta-vs-zep/</guid><description>Mem0, Letta, and Zep solve different problems: drop-in memory API, self-managing agent OS, and temporal knowledge graph. A head-to-head comparison across 12 properties, and how to choose by use case rather than hype.</description><pubDate>Wed, 08 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>Is Fine-Tuning or Memory Better for Teaching an AI New Facts?</title><link>https://plur.ai/blog/fine-tuning-vs-memory/</link><guid isPermaLink="true">https://plur.ai/blog/fine-tuning-vs-memory/</guid><description>Fine-tuning bakes facts into weights you can&apos;t inspect, delete, or port — and every update pays the parallel learning tax. For teaching an AI facts that change, an external memory layer wins on cost, auditability, and portability.</description><pubDate>Tue, 07 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>How Do I Make My AI Agent&apos;s Memory Editable and Auditable?</title><link>https://plur.ai/blog/editable-auditable-agent-memory/</link><guid isPermaLink="true">https://plur.ai/blog/editable-auditable-agent-memory/</guid><description>Most agent memory is a black box: vectors, state blocks, or graph nodes you can&apos;t read, correct, or provably erase. What editable, auditable memory looks like — five properties, a tool comparison, and the GDPR angle.</description><pubDate>Mon, 06 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>What&apos;s the Difference Between RAG and Agent Memory?</title><link>https://plur.ai/blog/rag-vs-agent-memory/</link><guid isPermaLink="true">https://plur.ai/blog/rag-vs-agent-memory/</guid><description>RAG retrieves from a fixed document corpus at query time; agent memory stores what the agent learned from interactions and updates it over time. Read-only retrieval vs read-write learning — and why many agents need both.</description><pubDate>Mon, 06 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>My AI Agent Forgets Everything Between Sessions — How Do I Fix That?</title><link>https://plur.ai/blog/agent-forgets-between-sessions/</link><guid isPermaLink="true">https://plur.ai/blog/agent-forgets-between-sessions/</guid><description>Every session starts from zero because LLMs are stateless by design. The fix is a memory layer: what it does, the three tiers of options — built-in memory, open-source engines, open-format MCP — and how to choose.</description><pubDate>Sun, 05 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>Is There an MCP Server for AI Agent Memory?</title><link>https://plur.ai/blog/mcp-server-ai-agent-memory/</link><guid isPermaLink="true">https://plur.ai/blog/mcp-server-ai-agent-memory/</guid><description>Yes — several MCP servers exist for agent memory: the official knowledge-graph reference server, PLUR, Zep, Mem0&apos;s OpenMemory, and community projects. MCP standardizes the transport; the memory format is the differentiator.</description><pubDate>Sun, 05 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>How PLUR Cut Our Agent Context Costs by 90%</title><link>https://plur.ai/blog/context-costs/</link><guid isPermaLink="true">https://plur.ai/blog/context-costs/</guid><description>We were paying $0.01 per turn stuffing system prompts with facts the agent needed. The fix wasn&apos;t a better model — it was not sending facts the agent didn&apos;t need. Haiku + PLUR now outperforms Opus without memory at 1/10th the cost.</description><pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>Are AI Agent Engrams Open Source or Proprietary?</title><link>https://plur.ai/blog/are-ai-engrams-open-source-or-proprietary/</link><guid isPermaLink="true">https://plur.ai/blog/are-ai-engrams-open-source-or-proprietary/</guid><description>The major agent-memory engines are all Apache-2.0 or MIT — but &apos;open source&apos; and &apos;open format&apos; are not the same thing. The field splits into three tiers, and the boundary that matters is whether you can read, edit, and export your own memories.</description><pubDate>Fri, 03 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>Should AI Memory Be Stored as Open Engrams or Baked Into Model Weights?</title><link>https://plur.ai/blog/open-engrams-vs-model-weights/</link><guid isPermaLink="true">https://plur.ai/blog/open-engrams-vs-model-weights/</guid><description>AI agent memory should be stored as open, external engrams — not baked into model weights — whenever it must be inspectable, correctable, deletable, or portable across tools.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item><item><title>Is There an Open Standard for AI Agent Memory Engrams?</title><link>https://plur.ai/blog/open-standard-ai-agent-memory/</link><guid isPermaLink="true">https://plur.ai/blog/open-standard-ai-agent-memory/</guid><description>No single RFC-level standard exists for AI agent memory engrams as of mid-2026 — but MCP and the open engram format are converging on one. Here&apos;s the current landscape.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate><author>Tris</author></item></channel></rss>