How it works Slack data question → warehouse answer and report.
Start a Qubot run A question in the Qubot Slack channel spawns a Copilot Cloud Agent instance on github.com[Source 1
Select context and query engine The run uses the federated context layer and routes work to Kusto for recent exploratory data or Trino for complex historical analysis[Source 1
Query the warehouse Qubot connects to Kusto and Trino through MCP servers and executes the data query[Source 1
Return answer and report The answer appears in Slack for threaded refinement and the results are stored as a Markdown report in a pull request[Source 1
Where people stay involved People review, approve, or guide the agent’s work here.
Not reported: The reviewed source shows threaded refinement but does not define a required human checkpoint.
Supervision level
Not classified
data question → warehouse answer[Source 1 How supervision levels work
Implementation details Model Powered by GitHub Copilot[Source 1 Harness Copilot Cloud Agent running on github.com[Source 1 Sandbox Not reported The source names the cloud-agent host but not its isolation boundary. Tool access Queries GitHub's data warehouse[Source 1 Knowledge Federated context contributed by teams and normalized into a structured context layer[Source 1 Context management A context agent ingests, organizes, and normalizes contributed context[Source 1 Credentials Not reported Interfaces Validation and failure handling Offline evaluation gate Context and agent-configuration changes run against curated prompts, ground-truth SQL, accuracy, latency, and regression checks before shipping[Source 1
Reported observations These results come from the sources. Notes below each result explain what it covers and any reported limits.
Headline claim Hundreds of users run thousands of queries; data questions in internal Slack channels dropped[Source 1
Reported by GitHub Scope Internal GitHub Qubot users and queries; no exact count or measurement window Observed in June 2026
About this result Category Adoption output Evidence type Qualitative What it covers Qubot adoption and data-channel question volume Key observation Hundreds of users run thousands of queries[Source 1
Reported by GitHub Scope Internal GitHub Qubot users and queries; no exact count or measurement window The source does not report the denominator of this figure.
About this result Category Adoption output Evidence type Estimate What it covers Internal Qubot users and queries Key observation Volume of data questions in internal data and analytics Slack channels dropped[Source 1
Questions in GitHub internal data and analytics Slack channels
About this result Category Adoption output Evidence type Qualitative What it covers Questions in internal data and analytics Slack channels Lessons Not reported: The reviewed source does not present a distinct transferable lesson in this record.
Back to agents Sources and research details Citations link to the original publisher. Each source also keeps a preserved copy in the repository, so a changed or removed page stays checkable.
How we built an internal data analytics agent https://github.blog/ai-and-ml/github-copilot/how-we-built-an-internal-data-analytics-agent/ Engineering blog · First party · Last source verification: 2026-08-31 Question coverage and scope purpose Reported workflow Reported human involvement Not reported: The reviewed source shows threaded refinement but does not define a required human checkpoint. implementation Reported validation Reported observations Reported lessons Not reported: The reviewed source does not present a distinct transferable lesson in this record. Research details for every claim on this page Summary Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Headline claim Statement type Metric Provenance Reported Confidence Medium Confidence reason GitHub reported the adoption figures in its own engineering blog without independent verification. Reported by GitHub Scope Internal GitHub Qubot users and queries; no exact count or measurement window Method Not reported Observation date 2026-06 Model Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Harness Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Tool access Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Interfaces Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Knowledge Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Context management Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Start a Qubot run Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Select context and query engine Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Query the warehouse Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Return answer and report Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Offline evaluation gate Statement type Fact Provenance Reported Confidence Not assessed Confidence reason No explicit confidence assessment is recorded for this claim. Key observation Statement type Metric Provenance Reported Confidence Medium Confidence reason GitHub reported the adoption figures in its own engineering blog. Reported by GitHub Scope Internal GitHub Qubot users and queries; no exact count or measurement window Denominator Not reported Method Not reported Observation date Not reported Key observation Statement type Fact Provenance Reported Confidence Medium Confidence reason Company reports a qualitative decrease without before-and-after counts. Scope Questions in GitHub internal data and analytics Slack channels Operating model assessment Statement type Inference Provenance Catalog judgment Confidence Unverified Confidence reason The public evidence does not document where human attention returns in the question-and-answer flow. Observation date 2026-06 