# Databricks — MemCurrent: Designing Persistent Memory for Production Agents

- Company: Databricks (databricks.com)
- Announced: 2026-10-20
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.databricks.com/blog/memcurrent-designing-persistent-memory-production-agents
- Record: https://forck.live/items/25339-memcurrent-designing-persistent-memory-for-production-agents
- Subject: Mosaic AI

Databricks introduced MemCurrent, a persistent memory architecture for production agents that now powers Agent Bricks Memory and Genie. The design combines three access patterns — plugin, agent tool, and offline pipeline — because each trades runtime cost and latency against how quickly updates take effect. MemCurrent keeps episodic memory of raw session traces and derived memory of distilled learnings in a database, with an auto-generated memory.md index and a dreaming pipeline that splits oversized files, merges duplicates and resolves conflicts. On Databricks' internal benchmark, built from real Genie conversations, MemCurrent triples the agent's pass rate over having no memory while using 31% fewer tokens and running 26% faster. The benchmark has 114 test cases, 29 users and 2,233 history sessions, drawn from 12K+ sessions and 3K mined candidates. Databricks says existing benchmarks are open-ended chats about simulated personas, while its own covers enterprise knowledge such as table semantics, metric definitions and workflows.

## Evidence

Verbatim from https://www.databricks.com/blog/memcurrent-designing-persistent-memory-production-agents:

> On our internal benchmark, built from realistic enterprise tasks, MemCurrent triples the agent's pass rate over having no memory, while using 31% fewer tokens and running 26% faster.

---

Record: https://forck.live/items/25339-memcurrent-designing-persistent-memory-for-production-agents
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md