live wire
▸Snowflake agent evals can pin judge versions; unpinned metrics may roll forward and shift scoresSnowflake▸ThoughtSpot makes Spotter 3 default for new instances; it requires download rights and always shares query data with an LLMThoughtSpot documentation▸MotherDuck adds image classification and object location to prompt_jev(), returning typed SQL outputs with confidence scoresMotherDuck▸BigQuery conversational analytics adds Montréal and Toronto; mixed-region agents still default to globalGoogle Cloud Documentation▸Fabric enables Azure OpenAI Copilot access by default, but OpenAI-subprocessor and cross-region processing/storage switches remain offMicrosoft Learn▸Snowflake Cortex Agents now continue by default when named tools are inaccessible, emitting warnings instead of failing the runSnowflake documentation▸Databricks Genie One can use group memberships to tailor answers, while Unity Catalog remains the access-control boundaryDatabricks documentation▸Fabric data agents can query Power BI models with Read alone; Build and workspace access are unnecessaryMicrosoft Learn▸Fabric data-agent Git configs expose column flags, but selecting a table still includes every columnMicrosoft Learn▸Google Data Agent Kit offers remote Spark kernels, but notebook exports can still move results into local Pandas or ArrowGoogle Cloud Documentation▸Cube says production analytics agents need explicit partial-result states, caller-filtered model search and durable artifactsCube▸Thinking Machines names Bridgewater behind ReViSQL-K2.6, citing 92.97% SC-16 accuracy at 15% of frontier-model costThinking Machines Lab▸AWS says a mismatched data/tool/prompt release cut its Field Sales agent pass rate from 78% to 38% in pre-productionAWS Artificial Intelligence Blog▸NL2SQL Multi-Agent ships evaluation artifacts across BIRD, CSpider, Archer and KDD DataSpace—but no software licenseGitHub▸Snowflake agent evals can pin judge versions; unpinned metrics may roll forward and shift scoresSnowflake▸ThoughtSpot makes Spotter 3 default for new instances; it requires download rights and always shares query data with an LLMThoughtSpot documentation▸MotherDuck adds image classification and object location to prompt_jev(), returning typed SQL outputs with confidence scoresMotherDuck▸BigQuery conversational analytics adds Montréal and Toronto; mixed-region agents still default to globalGoogle Cloud Documentation▸Fabric enables Azure OpenAI Copilot access by default, but OpenAI-subprocessor and cross-region processing/storage switches remain offMicrosoft Learn▸Snowflake Cortex Agents now continue by default when named tools are inaccessible, emitting warnings instead of failing the runSnowflake documentation▸Databricks Genie One can use group memberships to tailor answers, while Unity Catalog remains the access-control boundaryDatabricks documentation▸Fabric data agents can query Power BI models with Read alone; Build and workspace access are unnecessaryMicrosoft Learn▸Fabric data-agent Git configs expose column flags, but selecting a table still includes every columnMicrosoft Learn▸Google Data Agent Kit offers remote Spark kernels, but notebook exports can still move results into local Pandas or ArrowGoogle Cloud Documentation▸Cube says production analytics agents need explicit partial-result states, caller-filtered model search and durable artifactsCube▸Thinking Machines names Bridgewater behind ReViSQL-K2.6, citing 92.97% SC-16 accuracy at 15% of frontier-model costThinking Machines Lab▸AWS says a mismatched data/tool/prompt release cut its Field Sales agent pass rate from 78% to 38% in pre-productionAWS Artificial Intelligence Blog▸NL2SQL Multi-Agent ships evaluation artifacts across BIRD, CSpider, Archer and KDD DataSpace—but no software licenseGitHub
nl2sql.ai
releaseRELEASE

GoodData’s new report builder keeps AI out of the narrative—for now

The experimental feature turns live workspace data into branded PDF and PowerPoint deliverables, but explicitly stops short of autonomous reporting.

Live report versus exported files
Side by side: what changed
By The News Desk· Sep 18, 2026the quick take — two AI hosts go live when you do

GoodData added an experimental Reports surface on September 17, giving workspace users a presentation-style document builder for live analytics, reusable layouts and branded exports. The release is notable less for adding another dashboard format than for the boundary it draws: AI-generated narratives and conversational report editing are explicitly not part of this version. (GoodData release notes; Reports documentation)

A live object before export

A report combines saved visualizations, text, filters and a reporting period. GoodData describes it as a live workspace object: changing its date range re-executes the visualizations, while attribute filters can be applied across the report or configured per visualization. Templates are optional and copied when a report is created, so later template changes do not silently rewrite existing reports. (Reports documentation)

Branding is resolved from the workspace when the report is rendered rather than embedded in the report itself. The same report structure can therefore take on workspace-specific colors and assets. Dynamic text variables can record details including the reporting period, workspace, export time, exporting user and last editor. (Reports documentation)

The export is where the object becomes static. Users can generate PDF or PPTX files in the browser, and text remains editable in the PowerPoint output. An export does not freeze the source report; users can reopen it, change content or filters and create another snapshot. (Reports documentation)

The automation boundary is explicit

GoodData exposes workspace-scoped APIs to create, replace and delete reports, templates and page layouts. But those APIs cannot produce the final PDF or PPTX: exports require the browser flow, cannot run headlessly and cannot be scheduled unattended. The current release also lacks a visual template designer and separate view and edit modes. Report-specific access control and multi-tenant behavior are not finalized. (Reports documentation)

That distinction matters for teams assembling natural-language analytics workflows. This release can package governed, live workspace results into a reusable client-facing document, but it is not an autonomous reporting agent. A human still writes the narrative, initiates the export and operates inside an access model GoodData says is still being worked out. (Reports documentation)

The practical reading is to treat Reports as a human-reviewed delivery layer rather than a scheduled agent endpoint. Teams can automate creation and population through the API, but should not design production delivery around unattended exports or report-specific permissions until GoodData documents those capabilities. (Reports documentation)

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.