Graybeard

Docs

What it takes to run Graybeard, start to finish.

Getting started.

Graybeard is set up for you by the Graybeard team, not through a self-serve signup. Onboarding moves through scoping (your account, contacts, and the source documents you provide), provisioning, building your knowledge base, configuration, quality checks, and a live test on your surface before go-live. It is hands-on and measured in weeks. The engine is proven: the same generic pipeline has built knowledge bases for corpora as different as 554 products and six, with no one-off code for either.

A manufacturer's documentation being built into and operated as a running, searchable knowledge base.

API reference.

Partners can query Graybeard over a JSON API with an account key.

  • POST /v1/query sends a question and returns a cited answer
  • GET /v1/usage returns aggregate usage counters
  • Companion endpoints back the hosted embed.

Rate limits are per account: 500 requests per minute and 25,000 per day by default, overridable, with standard rate-limit headers. Keys and endpoints are issued by the Graybeard team at provisioning.

Putting Graybeard on your site.

Two paths, both set up per account by the team: a hosted window you embed, and the direct query API.

Managing your knowledge base.

When documentation changes, the team re-runs the build and the updated content replaces the old in place. Handled as part of your engagement, not a customer console.

Usage and reporting.

Graybeard reports what it is doing for you: how much question volume it is handling, which questions come up most, and which topics your documentation does not cover well yet, so you can see where a call was avoided and where a gap is worth closing.

Integrations.

Today Graybeard connects through the hosted embed or the API. Zendesk, Salesforce Knowledge, Freshdesk, HubSpot, Confluence, and custom systems are planned, and each gets setup steps here as it ships.

FAQ and troubleshooting.

Straight product answers: what Graybeard can ingest, how a citation is guaranteed to point at a real page, how the knowledge base is refreshed, how your data stays isolated, and the two ways to put Graybeard in front of customers. Customer questions from live use are added here as they come in.

What can Graybeard ingest?
The source documents you provide, including their text, tables, and diagrams.
How does Graybeard refresh its knowledge base?
When your documentation changes, the team re-runs the build and the updated content replaces the old in place.
What are the two ways to put Graybeard in front of customers?
A hosted window you embed, and the direct query API. Both are set up per account by the team.
Book a call