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

Progress expands Agentic RAG without a migration—but live-app access widens the review surface

The October release gives existing deployments a planning agent, a Teams interface and a source-side WordPress connector. The operational gain is real; so is the need to review every live system the agent can reach.

Existing Agentic RAG gains Smart Agent, Teams, and WordPress without migration.
Side by side: what changed
By The News Desk· Oct 3, 2026the quick take — two AI hosts go live when you do

Progress has shipped three capabilities to existing Agentic RAG deployments: a Smart Agent, a native Microsoft Teams app and a WordPress plugin. The company says customers receive them as configuration on their current platform, with no installation or migration required. That makes this more than a feature bundle: it is a test of whether an enterprise retrieval foundation can absorb new reasoning, delivery and ingestion paths without a rebuild. Progress

The agent now chooses when it has enough evidence

The Smart Agent decomposes a question into sub-questions, selects sources, re-queries when needed and evaluates whether the assembled answer is sufficient before stopping. Progress says quality scoring runs at each step, rather than only on the final response. It also says citations and audit logs preserve a trace of the agent’s steps for review. Progress

The consequential change is that retrieval no longer ends at indexed content. Through MCP connections, the agent can combine records from live applications such as CRM, ERP and ticketing systems with documents already in the knowledge base. Progress gives the example of joining an account’s CRM history, an open support ticket and a contract in one answer. Progress

That convenience also broadens the review surface. The launch post says steps are cited and audited, but it does not describe how permissions are reconciled when one answer spans several live systems. Teams access is clearer: Progress says the app carries over the deployment’s existing access controls and cited answers, without requiring a custom bot framework or authentication flow. Progress

WordPress moves freshness closer to the source

The WordPress plugin replaces a crawl-only path with a native, source-side integration intended to preserve structure and metadata while keeping content current. Progress characterizes it as a precision upgrade rather than a new source category. Progress

For practitioners, the release separates three deployment questions that are often bundled together: where an answer is delivered, which systems the agent may consult and how current the underlying content is. Existing customers avoid a migration, but adding live MCP applications should still trigger an access and audit review. The platform can ship configuration instantly; an organization’s authorization model still deserves deliberate validation.

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.