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
guideEnterprise operations

Managed databases break file-based recovery assumptions—IBM’s Db2 update shows the replacement

Db2 Recovery Expert 12.1 adds API-based archive-log access, KMIP support and tenant selection. For data-platform teams, the practical lesson is that managed-cloud recovery needs an explicit control path.

Db2 recovery path with managed-cloud branches and key management
AI-generated diagram
By The News Desk· Oct 4, 2026the quick take — two AI hosts go live when you do

IBM’s latest Db2 recovery release addresses a problem that managed databases make easy to overlook: a recovery tool cannot assume it can reach the database server’s file system.

Db2 Recovery Expert 12.1, generally available September 30, adds API-based reading of archive logs. IBM says this provides the basis for supporting Amazon RDS for Db2, where administrators do not have direct access to the underlying file system. A hybrid mode uses direct file access when available and switches to the API path when it is not. IBM announcement

Recovery now depends on the platform boundary

That change matters beyond one utility. Enterprise analytics and AI systems increasingly sit on managed databases, but their operational runbooks can still inherit assumptions from self-managed deployments: local archive logs, direct host access and one obvious database target. IBM’s release makes those assumptions explicit by adding separate mechanisms for managed-cloud log access, centralized encryption and tenant selection.

The release supports KMIP keystores used with Db2 native encryption. IBM says this removes the need to fall back to the Db2 instance keystore or manually migrate keys when an organization already uses centralized enterprise key management. It also adds namespace separation so an administrator can select the intended tenant connection before starting recovery. IBM announcement

Db2 pureScale environments get expanded remote log analysis, while custom VENDOROPT and LOGARCHOPT configurations are supported for teams using vendor-specific archive logging. IBM also aligned Recovery Expert’s versioning with Db2 and bundled IBM Semeru Runtime 25.0. IBM announcement

A deployment test worth adding

For teams operating analytics agents or natural-language query systems on top of Db2, this is not an argument that the agent should perform recovery. It is the opposite: query automation and recovery authority should remain separate, and the recovery path should be tested against the actual hosting model.

A practical acceptance test should therefore verify four things: archive logs remain readable without host access; encrypted logs can be opened through the organization’s approved key-management path; the operator must identify the intended tenant; and remote analysis works with the deployment’s archive-log configuration. Those checks follow directly from the boundaries IBM chose to address in 12.1. IBM announcement

The release does not claim to make recovery automatic or risk-free. What it does is replace hidden infrastructure assumptions with named interfaces—a necessary step when the database is managed by somebody else.

sources

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.

    more from the desk