Market Data Agent
Data marketing agent: answer a plain-English go-to-market question about companies and their technology installs (e.g. "US software companies with 5,000+ employees running Salesforce", "companies that replaced Splunk with a competitor", or "which companies are accelerating or declining on Snowflake"), the monthly install history of a product at a named set of companies or a segment (e.g. "monthly Snowflake detections at Fortune 500 companies"), and their estimated IT spend by category (e.g. a company's software spend, top spenders in Cloud Services), and technology contracts between customers and vendors (e.g. "which partners have the most SAP contracts", "Oracle contracts ending in the next 12 months"). It resolves the named entities, writes and EXECUTES a safe read-only query against the HG Insights data warehouse, and returns the result set — a summary and inline preview, plus a CSV artifact ("brief.csv") and a metadata sidecar ("metadata.json", with the executed SQL and schema) that you retrieve via phoenix_get_artifact. Asynchronous via polling: call with request to start; if the result is not ready in time it returns a task_id and status "working" — call the tool AGAIN with that task_id (omit request) to retrieve it. Out-of-scope asks (AI spend, spend over time, per-cloud spend, intent, contract renewal prediction, FAI) or unresolvable entities return a one-line decline instead of wrong data.
Credits
Metered separately — Billed through at cost via the underlying agent run. See the full credit table for how AI Credits work.
Parameters
| Name | Type | Default | Description |
|---|---|---|---|
request | string | - | Plain-English go-to-market question about companies, their technology installs and their contracts. Provide to START a query; omit when polling. |
task_id | string | - | Handle returned by a previous call. Provide to POLL for the result; omit to start a new query. |