Databricks put Agent Memory into Beta on Agent Bricks and made Unity AI Gateway generally available in October 2026
Databricks shipped the most agent-specific features of the month. Unity AI Gateway reached general availability on October 1, adding central governance, cost controls and observability across multiple models [68][69]. Agent Memory entered Beta on Agent Bricks later in the month. It stores facts, preferences and workflows in Lakebase and is built on the MemCurrent architecture, which pairs agent-written memory with an offline "dreaming" pipeline. Databricks reported that it tripled agent pass rate on its internal enterprise benchmarks [37][38][42]. Other releases included Lakebase Search, with vector and text extensions for agent retrieval [18], real-time Document Intelligence APIs for apps and agents [49], and Lakeflow Connect connectors for Atlassian, SharePoint, Google Workspace, GitHub and AI providers [15]. The agentic code converter became generally available for SQL [5]. CustomerLake's Agentic CDP entered Beta with Campaign Agents and Private Preview with Profile Agents [41].
Lakebase branching was a recurring theme for coding agents. Videos and posts showed agents working in parallel on isolated database branches [44][67], and Hamlet gave each pull request its own production-data branch [39]. Walkthroughs covered the Agent Bricks workflow with MLflow tracing and evaluation [43][53] and building an agentic app on Lakebase [52].
Outside Databricks, dbt Summit 2026 content dominated the news. Many posts argued that agents need governed business context: AWS traced a $3.5 million revenue error to missing semantic grounding [7], and Okta, Checkr, Kilo and Green Dot described their own context layers [8][16][23][26]. dbt Labs argued that agents could make the semantic layer succeed [13]. Others described shipping agents through dbt, including Careem, where about 90% of new models came from its dbt agent [20], and Under Armour [10].
Community discussion focused on agent memory [4][30][57], whether agentic analytics needs a separate semantic layer [1], and connection exhaustion in PostgreSQL caused by agents [50].
Everything cited
- [1]Does agentic analytics need a separate semantic layer if the data is already well modeled? community · 2026-10-31
- [2]Wealth Management Client Onboarding Needs More Than AI news · 2026-10-30
- [3]unit tests community · 2026-10-30
- [4]AI Agents Should Not Start Every Conversation From Zero community · 2026-10-30
- [5]Agentic code converter is GA for SQL community · 2026-10-29
- [6]LangGraph Tutorial - State, Nodes and Edges | Chapter 11 video · 2026-10-29
- [7]Why an AI agent's revenue query was $3.5 million wrong, and how AWS fixed it with dbt Semantic Layer news · 2026-10-27
- [8]How Checkr cut an hours-long metrics investigation to minutes by engineering AI context on dbt and Omni news · 2026-10-27
- [9]How LangChain grew self-serve analytics questions 40x by giving its AI agent layered context news · 2026-10-27
- [10]Ship AI agents like tables: dbt materializations for Snowflake CoWork news · 2026-10-27
- [11]Show HN: 10x faster, 3x cheaper ETL community · 2026-10-27
- [12]With great context comes great autonomy news · 2026-10-27
- [13]Why dbt Labs thinks AI agents will finally make the semantic layer stick news · 2026-10-27
- [14]Do any other teams build web apps for fully flexible analytics/reporting? community · 2026-10-27
- [15]Unify your productivity and AI data with Lakeflow Connect news · 2026-10-26
- [16]What Okta learned governing context for AI agents at scale news · 2026-10-26
- [17]Inside dbt Wizard, the AI agent built to show its work news · 2026-10-26
- [18]Lakebase Search: Build AI Retrieval for Agents with Lakebase Postgres video · 2026-10-26
- [19]Governing AI Agents with Unity Catalog and Unity Gateway on Databricks | Unity Catalog Demo video · 2026-10-26
- [20]From prompt to PR: how Careem’s AI writes and validates dbt models before a human ever sees them news · 2026-10-26
- [21]How Ramp's internal AI agents crossed a million sessions news · 2026-10-26
- [22]What a swarm of rogue AI agents can teach us about having better ideas news · 2026-10-26
- [23]How a data scientist at Kilo built the business context layer his AI agents were missing news · 2026-10-26
- [24]How to solve dashboard debugging in six minutes with an MCP-connected AI agent news · 2026-10-26
- [25]When skills meet scale: an agent assembly line for dbt news · 2026-10-26
- [26]Meta:Context: a business context schema in dbt's semantic layer news · 2026-10-26
- [27]The media industry is becoming a prediction economy news · 2026-10-26
- [28]Enterprise Data with AI Agents: A Data Platform Guide community · 2026-10-26
- [29]Test on a Branch, Not on Patients: Agentic Drug Substitution with Lakebase community · 2026-10-25
- [30]AI Agent Memory community · 2026-10-25
- [31]Agent Bricks: Production Ready AI Agents on your Enterprise Data community · 2026-10-23
- [32]Can AI Really Model Your Data in Hours? Databricks Vibe Modeling video · 2026-10-22
- [33]The Fraud Analyst AI Coworker: Detecting Telecom Fraud in Real-Time on One Platform news · 2026-10-22
- [34]Building Reliable Data Pipelines with Lakeflow on Databricks | Lakeflow Demo video · 2026-10-22
- [35]Building an Agentic Analytics Foundation on the Lakehouse video · 2026-10-21
- [36]Free live session on Kafka as the event log for AI agent memory community · 2026-10-20
- [37]MemCurrent: Designing Persistent Memory for Production Agents news · 2026-10-20
- [38]Introducing Agent Memory on Databricks news · 2026-10-20
- [39]How Hamlet uses Databricks Lakebase Postgres to test changes against production data news · 2026-10-20
- [40]How to Build Customer 360 Profiles with AI | CustomerLake Profile Agents Demo video · 2026-10-19
- [41]CustomerLake enters Beta: Leading enterprises going live with the Agentic CDP news · 2026-10-19
- [42]Introducing Agent Memory on Databricks video · 2026-10-19
- [43]Building Production-Ready AI Agents on Databricks | Agent Bricks Demo video · 2026-10-19
- [44]Shorts extracted from Parallel Coding Agents with Lakebase video · 2026-10-19
- [45]Databricks Is Doubling Down On New York City As A Premier AI Talent Hub news · 2026-10-19
- [46]Why Halliburton's SVP of Technology Says Don't Start With AI | Sriram Srinivasan, Halliburton video · 2026-10-19
- [47]From Excel to AI Agents: The Evolution of BI Explained community · 2026-10-16
- [48]Demo: Intelligent Document Processing in Databricks | Feat. Databricks' Sepi video · 2026-10-16
- [49]Power real-time applications and agents with Document Intelligence news · 2026-10-14
- [50]Why AI agents cause connection exhaustion in PostgreSQL (and how we fixed it with 17 connections) community · 2026-10-14
- [51]Announcement | Lakebase and Agentic SDLC: Branching Databases for Coding Agents community · 2026-10-14
- [52]Building on Lakebase | Episode 2: Anatomy of an AI app: build the operational + agentic app video · 2026-10-13
- [53]Agent Bricks: The Agent Developer Platform | Overview video · 2026-10-12
- [54]Single DAB covers your entire agentic app architecture on Databricks community · 2026-10-12
- [55]How HelloFresh Optimizes 100% of Meta Ad Spend With Genie Agents on Databricks video · 2026-10-09
- [56]Run Google Dataproc (Managed Spark) on local, for offline dev, CI, and AI agent testing (multi-arch, zero cloud cost) community · 2026-10-09
- [57]Beyond Chat History: Designing Portable Memory and Decision Boundaries for Enterprise AI Agents community · 2026-10-09
- [58]Building on Lakebase | Episode 1: Meet Lakebase video · 2026-10-06
- [59]Recommendations for limiting spend on a pay-as-you-go account community · 2026-10-06
- [60]A song about Fabric & Databricks :) community · 2026-10-05
- [61]Virtual Event | The Emerging Blueprint for Agentic Apps community · 2026-10-05
- [62]How to choose the right AI agent path on Databricks community · 2026-10-05
- [63]Data Lakehouse with Agentic AIs: A Guide community · 2026-10-05
- [64]Data Lakehouse with Agentic AIs: A Guide community · 2026-10-05
- [65]I built an open-source check that stops AI agents from inventing column names in SQL. community · 2026-10-04
- [66]Building A Data App with dashboard and agentic capabilities in dbx community · 2026-10-04
- [67]Parallel Coding Agents with Lakebase | Claude Code + GitHub Actions video · 2026-10-01
- [68]How Enterprises Govern AI Agents Across Multiple Models video · 2026-10-01
- [69]Demo: Building a Governed AI Agent with Unity AI Gateway video · 2026-10-01
- [70]Stellantis: How a Global Automaker Deployed AI Agents at Scale video · 2026-10-01
- [71]How ModMed Transforms Healthcare AI and Agentic Workflows with Databricks video · 2026-10-01
A frozen monthly snapshot, generated from the brickster.ai archive and never rewritten. For the live view of this topic, see the AI Agents hub. brickster.ai is an independent community project, not affiliated with Databricks, Inc.