Databricks unveils MemCurrent, a persistent memory architecture for AI agents
What happened
On October 20, Databricks detailed MemCurrent, a production-grade persistent memory architecture for AI agents. Reports say it is already used in Agent Bricks Memory and Genie. MemCurrent combines episodic memory that keeps raw sessions with derived memory that distills experience, reading, writing and organizing memory through plugins, on-demand tools and an offline dreaming pipeline. On an internal benchmark of real enterprise tasks, pass rate rose from 22.3% without memory to 68.4%, token use fell 31% and runs were 26% faster.
Written by AI from the coverage · updated 1 hour ago
Coverage
Follow the reports to see the story from different sides.
- AI HOT · ProductsDatabricks 发布 MemCurrent 生产级 AI 智能体持久记忆架构
Databricks 推出 MemCurrent 持久记忆架构，已用于 Agent Bricks Memory 和 Genie。它结合保存原始会话的情景记忆与提炼经验的派生记忆，通过插件、按需工具和离线 dreaming 流水线读取、写入及整理记忆。在来自真实企业任务的内部基准上，相比无记忆方案，通过率从 22.3% 提升至 68.4%，token 用量减少 31%，运行速度加快 26%。
Heat over time
Not enough continuous observations to draw a trend yet.