live wire
IBM makes watsonx Orchestrate AgentOps, custom LLM judging and Bedrock-agent discovery generally availableIBMSchemaGate 0.1.45 fixes broken Oracle ADB wallet connections and an OCI stack pinned 28 releases behindSchemaGatePDI’s Amazon Quick procurement agent grounds spend answers in vendor, category and contract contextAWS Business Intelligence BlogBigQuery’s ML.METRICS example returns 0.84 accuracy but 0.30 macro-F1 on the same 100-row classification queryGoogle Cloud BigQuery docsSchemaGate 0.1.44 auto-selects sentence embeddings, lifting bundled-schema retrieval from 90/98 to 93/98SchemaGateSchemaGate 0.1.43 adds read-only SQL execution with per-principal table checks—and documents unauthenticated client assertionsSchemaGateDatabox adds reusable AI Analyst Skills with personal/company scope, auto-matching and marketplace installsDataboxFabric previews an AI builder for data-agent instructions, source guidance and example queriesMicrosoft FabricDatabricks trains data-agent retriever to stop early or spend bounded extra search steps, reporting 5.8-second latencyDatabricksThoughtSpot adds SpotterCode coding agent to its Visual Embed PlaygroundThoughtSpotLongMemEval-S audit: 67–73% of restore-fixable 80k-budget errors came from evicted evidence under three policiesarXivSchemaGate 0.1.42 adds dimension-aware retrieval and fixes complex multi-table SQL promptsSchemaGateSnowflake agent toolsets can silently drop inherited tools when callers lack accessSnowflake DocumentationLooker’s VS Code extension reaches GA with MCP-assisted LookML generation, editing and validationGoogle Cloud Looker release docsIBM makes watsonx Orchestrate AgentOps, custom LLM judging and Bedrock-agent discovery generally availableIBMSchemaGate 0.1.45 fixes broken Oracle ADB wallet connections and an OCI stack pinned 28 releases behindSchemaGatePDI’s Amazon Quick procurement agent grounds spend answers in vendor, category and contract contextAWS Business Intelligence BlogBigQuery’s ML.METRICS example returns 0.84 accuracy but 0.30 macro-F1 on the same 100-row classification queryGoogle Cloud BigQuery docsSchemaGate 0.1.44 auto-selects sentence embeddings, lifting bundled-schema retrieval from 90/98 to 93/98SchemaGateSchemaGate 0.1.43 adds read-only SQL execution with per-principal table checks—and documents unauthenticated client assertionsSchemaGateDatabox adds reusable AI Analyst Skills with personal/company scope, auto-matching and marketplace installsDataboxFabric previews an AI builder for data-agent instructions, source guidance and example queriesMicrosoft FabricDatabricks trains data-agent retriever to stop early or spend bounded extra search steps, reporting 5.8-second latencyDatabricksThoughtSpot adds SpotterCode coding agent to its Visual Embed PlaygroundThoughtSpotLongMemEval-S audit: 67–73% of restore-fixable 80k-budget errors came from evicted evidence under three policiesarXivSchemaGate 0.1.42 adds dimension-aware retrieval and fixes complex multi-table SQL promptsSchemaGateSnowflake agent toolsets can silently drop inherited tools when callers lack accessSnowflake DocumentationLooker’s VS Code extension reaches GA with MCP-assisted LookML generation, editing and validationGoogle Cloud Looker release docs
nl2sql.ai
releaseRelease

ThoughtSpot 26.8.0 makes agent rollout a scoping and control problem

The release couples specialized Spotter configurations with connector testing, local-time interpretation and separate raw-data export privileges.

Scoped analytics agent controls versus broad access.
Side by side: what changed
By The News Desk· Aug 28, 2026the quick take — two AI hosts go live when you do

ThoughtSpot’s Cloud 26.8.0 release is less about giving every user a larger general-purpose assistant than about narrowing what an analytics agent is allowed to know and do. Its new Spotter Analysts, currently labeled early access, let administrators create focused analysts around specified data models, custom instructions and connectors. That makes the agent’s operating context an explicit configuration rather than an implicit consequence of whatever content happens to be available. ThoughtSpot’s release page lists no general-availability date for the feature.

Scope first, then test the edges

The same release adds a direct test for MCP connectors inside Spotter’s AI settings. ThoughtSpot describes it as a way to catch connector errors before users encounter them. In practice, that creates a preflight step for one of the riskiest boundaries in an analytics-agent deployment: the handoff from a language-model interaction to an external tool or data service. The release also adds Spotter Starter Prompts, which administrators can use to steer users toward supported searches and analyses rather than leaving every session at a blank prompt. Both controls appear in the 26.8.0 feature list.

A separate Timezone Aware Keyword feature addresses a deceptively common source of wrong answers. Relative terms such as “today” and “month-to-date” can now follow local business hours instead of a global default. For teams comparing regional operations, that is not interface polish: it determines which records fall inside a requested period. ThoughtSpot explicitly ties the feature to local business hours.

Separate insight sharing from data extraction

ThoughtSpot also splits download privileges for visual exports, such as PDF and PNG, from raw-data exports in XLSX and CSV. That gives administrators a finer distinction between letting someone circulate an insight and letting them extract its underlying rows. The release’s early-access Share Spotter Chat feature keeps shared conversations inside ThoughtSpot and describes them as governed, while Insight Tiles, also early access, run real-time code to generate live takeaways in Liveboards. The company lists all three capabilities in the same release.

The release rounds out the operating model with SpotterCode authentication for live code execution and cluster administration through REST APIs, plus a native ClickHouse connector. The important pattern is the combination: specialized agent scope, connector preflight, time semantics, execution credentials and separate export permissions. Buyers evaluating 26.8.0 should test those boundaries as one control chain, not as unrelated feature checkboxes. ThoughtSpot’s primary release page is the source for the feature set.

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.