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
newsEnterprise launch

Setplex puts read-only conversational analytics inside its video operations platform

AI Insider answers natural-language questions over live service, audience and content data while keeping model access isolated from subscriber records.

By The News Desk· Sep 12, 2026the quick take — two AI hosts go live when you do

Setplex has launched AI Insider, a conversational analytics agent built into the analytics layer of its Zapflex video service platform. The company says operators, broadcasters and media owners can ask natural-language questions about service quality, content performance and audience activity, then receive relevant metrics, charts and visual answers. Setplex announced the product on September 10.

The design is aimed at questions that do not fit a predefined dashboard. Setplex says users can start with a broad question, refine it through follow-up prompts and keep the context of the conversation. A feature called Smart Chip proposes drill-downs, comparisons with other periods and adjacent metrics based on the active question rather than simply replaying query history. The launch post describes those conversational and recommendation behaviors.

The safety boundary is implemented in code

The most consequential part of the launch is its execution model. Setplex says AI Insider runs through an isolated workflow that prevents the model from directly accessing subscriber data. Requests pass through controlled, verifiable nodes, and read-only access is enforced in code rather than left to prompt instructions. The company also says operators do not need to export their data, deploy a separate BI system or place a model inside the database. Those controls are stated in Setplex’s product announcement.

That boundary matters operationally because conversational analytics combines two different risks: exposing sensitive context to a model and allowing generated requests to change production state. Setplex’s architecture addresses those risks separately—isolating subscriber data from direct model access and making the analytics path read-only. Its announcement does not publish an independent security assessment, so buyers still need to verify the implementation in their own review.

Setplex says every prompt and generated query is logged, creating an audit trail for governance and troubleshooting. It also includes rate limits and cost ceilings intended to bound usage and spending. The vendor presents those controls as part of making the agent suitable for production.

What teams should test

A pilot should verify that read-only enforcement holds across every query path, that logs connect each prompt to the generated query and returned result, and that cost ceilings fail closed when reached. Teams should also test whether conversational follow-ups preserve the intended time range and population as questions become more specific. Those checks follow directly from the product’s claimed isolation, audit and multi-turn behavior.

Setplex says it serves more than 80 operators across five continents and will demonstrate AI Insider at IBC2026 in Amsterdam. The announcement provides the company and event details. The launch adds a vertical-specific example of conversational analytics moving from a general BI surface into the operational platform where service decisions are made.

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