live wire
▸Databricks will default Genie One on for account-only users; objects shared to “all account users” need reviewDatabricks▸Databricks makes Genie Code CLI GA, but disabling its web search does not stop shell-based internet accessDatabricks▸OpenData Agent finds 3B single-agent and router-specialist designs nearly tied on 120 French RAG+SQL questionsOpenData Agent repository▸Databricks says ai_decide is 33% more accurate, while its beta contract permits the underlying model to changeDatabricks▸Adaptive-Elbow reaches 97% table recall at 36.9% precision on merged BIRD Mini-Dev, but does not test SQL executionAdaptive-Elbow paper▸Databricks Genie One scheduled tasks can now email other users, including visualizations and a PDF of each runDatabricks▸ThoughtSpot Data Cache is GA for Snowflake and Databricks models, but cached answers can differ from live resultsThoughtSpot▸Tableau raises Cloud Manager ceiling to 200 sites, but Cloud+/Tableau+ base capacity still supports 50Tableau▸Databricks adds daily per-user Genie spending controls, but blocking remains approximate and uses list pricesDatabricks▸Databricks brings governed-tag propagation to Runtime 19+ classic compute, but ABAC can lag derived data by minutesDatabricks▸Snowflake adds Pegasus 1.5 video analysis to AI_COMPLETE at GA, while video processing remains public previewSnowflake▸Living Context Graph raises held-out BIRD transfer accuracy from 13.7% to 36.3%, but needs verified SQL outcomesLiving Context Graph paper▸Google previews Apache Iceberg views in BigQuery’s Lakehouse catalog, but query-time access still reaches every base tableGoogle Cloud▸Oracle makes OCI Agent Sandbox GA with managed Linux execution, project-scoped egress and 48-hour initial lifetimesOracle OCI▸Databricks will default Genie One on for account-only users; objects shared to “all account users” need reviewDatabricks▸Databricks makes Genie Code CLI GA, but disabling its web search does not stop shell-based internet accessDatabricks▸OpenData Agent finds 3B single-agent and router-specialist designs nearly tied on 120 French RAG+SQL questionsOpenData Agent repository▸Databricks says ai_decide is 33% more accurate, while its beta contract permits the underlying model to changeDatabricks▸Adaptive-Elbow reaches 97% table recall at 36.9% precision on merged BIRD Mini-Dev, but does not test SQL executionAdaptive-Elbow paper▸Databricks Genie One scheduled tasks can now email other users, including visualizations and a PDF of each runDatabricks▸ThoughtSpot Data Cache is GA for Snowflake and Databricks models, but cached answers can differ from live resultsThoughtSpot▸Tableau raises Cloud Manager ceiling to 200 sites, but Cloud+/Tableau+ base capacity still supports 50Tableau▸Databricks adds daily per-user Genie spending controls, but blocking remains approximate and uses list pricesDatabricks▸Databricks brings governed-tag propagation to Runtime 19+ classic compute, but ABAC can lag derived data by minutesDatabricks▸Snowflake adds Pegasus 1.5 video analysis to AI_COMPLETE at GA, while video processing remains public previewSnowflake▸Living Context Graph raises held-out BIRD transfer accuracy from 13.7% to 36.3%, but needs verified SQL outcomesLiving Context Graph paper▸Google previews Apache Iceberg views in BigQuery’s Lakehouse catalog, but query-time access still reaches every base tableGoogle Cloud▸Oracle makes OCI Agent Sandbox GA with managed Linux execution, project-scoped egress and 48-hour initial lifetimesOracle OCI
nl2sql.ai
guideROLLOUT GUIDE

Genie One scheduled tasks can now email colleagues—treat the recipient list as a data-delivery surface

Databricks expanded recurring Genie outputs beyond the task owner. The useful rollout question is no longer only whether the prompt is correct, but who receives the rendered result and attachment.

By The News Desk· Nov 17, 2026the quick take — two AI hosts go live when you do

Databricks has expanded Genie One scheduled tasks so they can email other users in the same organization, rather than sending results only to the task author. The November 12 release note makes the recipient change explicit; the product documentation says each run posts into a chat thread and sends an email containing the results, any visualizations, and a PDF attachment of the output. Databricks release notes Scheduled-task documentation

That turns a personal automation feature into a lightweight recurring distribution channel. A user can create a task in natural language or through a manual form, where Databricks exposes fields for the title, instructions, connections, schedule and timezone. Authors can also run the task immediately, inspect past runs, edit it, delete it or mention it from a new request. Scheduled-task documentation

The new control point is the rendered output

The operational unit to review is not just the prompt. It is the email package: answer text, visualizations and an attached PDF. Those artifacts can be forwarded, retained or downloaded outside the original chat thread. Databricks documents the package and the new ability to address other organizational users, but the cited scheduled-task section does not describe a recipient-specific source-permission recheck. Teams should therefore test the behavior with representative recipient accounts before using the feature for sensitive recurring reports. Scheduled-task documentation

A sensible rollout starts with three checks:

  • Use a low-sensitivity report first. Validate the exact email body, visualization content and PDF before adding a broader recipient list.
  • Review the configured connections. The manual task form includes a Connections field, so the delivery review should cover every source the scheduled instruction can reach.
  • Re-test after edits. Use Run now after changing instructions, connections or recipients, and retain the resulting thread as the review record.

These are rollout recommendations, not controls Databricks says it enforces automatically.

Why this small release matters

Scheduled analytics often become de facto reporting infrastructure. Genie One now supplies the creation interface, execution history, recurring cadence and outbound email package in one workflow. That is convenient, but it also means a seemingly minor recipient edit can change the audience for a generated analysis and its PDF copy. Scheduled-task documentation

Administrators do not need to block the feature to manage that risk. They do need to inventory shared schedules, assign an owner, review recipient lists and treat output previews as part of the approval process. The release is most useful when teams pair its convenience with the same distribution discipline they already apply to emailed dashboards and scheduled BI exports.

Filed by The News Desk. Corrections: desk@nl2sql.ai · Our standards →

comments · 0

    Comments are moderated before they appear. Your email is used once to confirm it is you — never shown, never sold. Corrections and questions get an answer from the desk when we have one.