The Chat API lets your application submit a natural-language question and get a Databook answer immediately. It's the same conversational engine that powers DatabookAI in the Databook app, made available wherever you need it.
Use the Chat API when a person or an agent is waiting on the answer. If you need to run insights on a schedule across many accounts, use the Batch API instead.
What it can do
Answer any question DatabookAI can answer: account intelligence, financial performance, strategic priorities, business and financial challenges, competitive landscape, buyer and contact strategy, discovery questions, meeting preparation, and account planning
Apply full DatabookAI reasoning: your response uses the same prompt planning as the Databook app, so answers are longer and more contextual than Batch API output
Run coaches and produce assets: the Chat API can run Databook coaches and return the assets they produce
Hold a conversation: a conversation identifier is returned with each answer, and you can pass it back on your next request so follow-up questions keep their context
Accept custom prompting: you control the wording of the question, so you can shape tone, length, and output format
How it works
Authenticate to obtain an access token.
Send a natural-language query to the chat endpoint, optionally including a conversation identifier from an earlier exchange.
Databook resolves the company in your question using its entity linker, retrieves the relevant material, and generates the answer.
The response comes back as JSON containing the answer text and a conversation identifier for follow-ups.
Company matching uses the same entity resolution as the Databook app. A question can resolve to a parent company or a subsidiary, depending on which has the richer insight coverage and the better contextual fit for your question.
Example use cases
Chat tools: ask a question from inside Slack or Teams and get the same answers, coaches, and assets you'd get in the Databook app
Embedded insight: surface real-time insights inside your own application, next to the record you're already working on
Agents and copilots: build an agent or copilot that calls Databook when it needs account intelligence
Custom front ends: build a purpose-built seller experience on top of Databook's capabilities
A representative query looks like: "What are the latest business challenges for Walmart?"
What to expect from a response
Answers are generated per request from Databook's premium data sources and live retrieval, so the same question asked twice may not return identical wording
Response times vary with the complexity of your question, since some questions trigger additional retrieval such as web search
Citation and disclaimer requirements apply wherever the output is displayed
Limits worth knowing
Limit | Value |
Maximum query length | 20,000 characters |
Request timeout | 180 seconds |
Rate limiting | Applied to keep the service reliable |
Streaming responses | Not available, responses are returned in full |
Getting access
Chat API access requires an entitlement and credentials issued by Databook. Once credentials are in place, no further configuration is needed: any question you can ask in the Databook app, you can ask through the API. Full request and response detail is in the API documentation portal.