Skip to main content
Driven connects Agents to professional financial data sources for market research and analysis. This is the foundation that separates Driven’s answers from a general model’s: figures come from databases, not from the model’s memory. When you ask for a margin or a holding, the Agent pulls it from a live source rather than recalling a number that may be stale or invented. Driven’s coverage spans 245 data endpoints across 28 categories. It reaches across US, Hong Kong, and A-share equities, plus crypto, FX, commodities, and macro time series.

​
Data categories

The 28 categories span the full research workflow. Coverage can include:
  • Real-time and historical quotes
  • Financial statements
  • SEC and exchange filings
  • Earnings and dividend calendars
  • Institutional holdings
  • Insider activity
  • Technical indicators
  • Market news
  • ETF and fund data
  • Crypto, FX, and commodities
  • Macro time series

​
Signature data

Beyond standard market data, Driven exposes a set of signature datasets that most general tools miss:
  • Congressional trades — disclosed trading activity by members of Congress
  • 13F institutional holdings — quarterly positions reported by institutional managers
  • Insider Form 4 filings — insider buys and sells filed with the SEC
  • Earnings-call transcripts — full transcripts of company earnings calls
  • SEC 8-K filings — material event disclosures
  • ETF reverse-lookup — find which ETFs hold a given stock
  • Analyst price targets — published target prices and revisions
  • Community pulse — social and community sentiment from sources like Xueqiu and X

​
Hong Kong, A-share, and China ADR data

Financial statement data for Hong Kong, A-share, and Chinese ADR stocks is sourced to match local accounting standards, with faster refresh cycles and cleaner field mappings. On top of standard statements, Driven covers three categories of A-share-specific data that general tools usually miss:
  • Capital flow — institutional versus retail flow, Dragon & Tiger (龙虎榜) lists, block trades, and margin trading (融资融券)
  • Shareholder structure — shareholder count, share capital breakdown, institutional holdings and their changes, and executive share changes
  • Company intelligence — thematic sector links (概念联动), business segment commentary, share pledging (股权质押), valuation bands, historical earnings, and peer comparison

​
Research coverage vs trading coverage

Research coverage can be broader than paper trading coverage. You can often research a market that you cannot yet paper trade. Before building an automated workflow, confirm that the market, asset type, and account type you need are all supported.

​
Best practice: ask the Agent to cite its data

For investment work, always have the Agent show its sources. This makes the analysis auditable and surfaces any gaps: 
Analyze NVDA and cite the financial statement fields, filings, and market data used. Flag any unavailable data.
If the Agent could not retrieve something, a good prompt makes it say so rather than filling the gap with a guess.