live wire
▸Snowflake’s Cortex Agent quickstart routes metrics and transcripts with prompt rules—but documents no authorization boundarySnowflake▸Oracle Select AI generates SQL/PGQ for property graphs—but errors if one AI profile mixes graphs with tables or viewsOracle▸Domo’s data team standardizes MCP builds with shared scaffolding; beta toolkits inherit user permissionsDomo▸Databricks proposes five outcome signals for agentic BI—and says dashboard counts and license usage miss decision qualityDatabricks▸One vibe-coded database app needed 56+ human test hours and three repair rounds despite AI Studio’s verification claimsMancas & Mancas / arXiv▸BigQuery AI.IF’s default cost mode trains a local model for most rows—and drops row-error thresholdsGoogle Cloud▸AWS’s multi-tenant data-agent pattern gives the broker Lambda no data permissions; Cognito roles carry Lake Formation scopeAWS Big Data Blog▸Amazon OpenSearch’s optimized engine adds nested agent traces, but AWS reserves it for log analytics—not vector searchAWS▸Two interpretation-based provenance semantics cannot express SQL bag difference, new theorem showsKolaitis / arXiv▸Databricks Feature Views reuse one definition for training and live scoring, but online lookup still returns latest available stateDatabricks▸Databricks Genie’s Slack and Teams previews default channel answers to public, even when viewers lack source-data accessDatabricks▸Snowflake’s two-tool GraphRAG agent beat a seven-tool graph setup 70% to 60% on 22 ontology questionsSnowflake Engineering▸Semantix opens an Apache-2.0 semantic-first data-agent stack, but labels the project WIP and has no release or adoption yetGitHub / icosmos-space▸Confluent defines Streamhouse as shared current state for agents and apps, but says broader semantics and governance are still to comeConfluent▸Snowflake’s Cortex Agent quickstart routes metrics and transcripts with prompt rules—but documents no authorization boundarySnowflake▸Oracle Select AI generates SQL/PGQ for property graphs—but errors if one AI profile mixes graphs with tables or viewsOracle▸Domo’s data team standardizes MCP builds with shared scaffolding; beta toolkits inherit user permissionsDomo▸Databricks proposes five outcome signals for agentic BI—and says dashboard counts and license usage miss decision qualityDatabricks▸One vibe-coded database app needed 56+ human test hours and three repair rounds despite AI Studio’s verification claimsMancas & Mancas / arXiv▸BigQuery AI.IF’s default cost mode trains a local model for most rows—and drops row-error thresholdsGoogle Cloud▸AWS’s multi-tenant data-agent pattern gives the broker Lambda no data permissions; Cognito roles carry Lake Formation scopeAWS Big Data Blog▸Amazon OpenSearch’s optimized engine adds nested agent traces, but AWS reserves it for log analytics—not vector searchAWS▸Two interpretation-based provenance semantics cannot express SQL bag difference, new theorem showsKolaitis / arXiv▸Databricks Feature Views reuse one definition for training and live scoring, but online lookup still returns latest available stateDatabricks▸Databricks Genie’s Slack and Teams previews default channel answers to public, even when viewers lack source-data accessDatabricks▸Snowflake’s two-tool GraphRAG agent beat a seven-tool graph setup 70% to 60% on 22 ontology questionsSnowflake Engineering▸Semantix opens an Apache-2.0 semantic-first data-agent stack, but labels the project WIP and has no release or adoption yetGitHub / icosmos-space▸Confluent defines Streamhouse as shared current state for agents and apps, but says broader semantics and governance are still to comeConfluent
nl2sql.ai
guideOracle Select AI

Oracle Select AI can query property graphs—but not from the same profile as tables

The new SQL/PGQ path is useful for relationship-heavy questions, but Oracle’s object-list rule creates a hard architecture boundary for mixed relational-and-graph assistants.

Oracle Select AI must separate graph and table queries.
Side by side: what changed
By The News Desk· Nov 8, 2026the quick take — two AI hosts go live when you do

Oracle’s November 2026 Select AI guide adds a useful new target for natural-language database questions: SQL Property Graphs. The important deployment detail is not just that Select AI can generate graph queries. It is that Oracle requires graph context to live apart from relational context inside an AI profile.

How the graph path works

When an AI profile’s object_list contains a property graph, Select AI sends the model instructions for constructing Property Graph Query (PGQ) plus metadata derived from the graph’s CREATE PROPERTY GRAPH statement. It then generates a query through Oracle’s GRAPH_TABLE operator, runs it and returns the result. Oracle says that syntax belongs to the SQL/PGQ section of the SQL:2023 standard.

The same interface still routes ordinary tables, views and other relational objects to SQL. For graph prompts, supported actions include runsql, showsql, explainsql and narrate; the feature can also retain short-term session context or customizable long-term conversation context.

The profile boundary matters

Oracle explicitly says that a profile cannot mix a property graph with tables, schemas or views. If it does, Select AI raises an error. This means a production assistant that needs both conventional business metrics and graph traversal cannot simply place every object in one Select AI profile and let the model choose.

The safer design is to keep a graph-only profile and a relational profile, then route the user’s request before invoking Select AI. That router becomes part of the application’s correctness boundary: relationship questions can go to the graph profile, while aggregate and row-oriented questions can go to the relational profile. Cross-domain questions need an application-level composition step rather than an assumed mixed query.

Know what is not supported

The feature supports SQL Property Graphs, not Oracle’s older PGQL Property Graphs. Oracle also lists synthetic-data generation and mixed graph-and-relational profiles as unsupported. Its documentation flags some queries as unsupported or intermittent, so teams should evaluate their own graph-question set rather than treating natural-language support as universal coverage.

This release broadens Select AI beyond familiar NL2SQL over tables, but it does not erase data-model boundaries. For practitioners, the operational takeaway is straightforward: graph metadata can now ground natural-language PGQ generation, while routing and result composition remain application responsibilities.

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