Skip to main content
Vizkraft helps your team explore data in plain language, turn answers into charts, and assemble dashboards in minutes. You can connect managed databases, HubSpot CRM, or custom MCP servers, ask questions across one or more connectors, save charts, apply filters, teach Vizkraft with connector memory, and keep AI-suggested content that is generated after your data is indexed.

​
What you can do in Vizkraft

  • Add connectors for PostgreSQL, MySQL, ClickHouse, MongoDB, HubSpot CRM, and custom MCP servers.
  • Index SQL and MongoDB sources, or sync HubSpot CRM catalogs.
  • Ask questions in chat and keep the conversation going with follow-up prompts.
  • Rate answers with thumbs up or thumbs down to improve accuracy over time.
  • Build connector memory that persists across your organization.
  • Run evaluation to measure and improve answer quality for SQL connectors.
  • Create charts from answers, save them, refine them manually or with AI edit.
  • Build dashboards with manual filters, AI filters, smart filters, saved views, and exports.
  • Review suggested charts and dashboards that Vizkraft prepares after a connector is analyzed.
  • Work from the web app and, if enabled for your workspace, from Slack.

​
Your main workspace areas

AreaWhat it is for
ChatAsk questions, review answers, rate responses, and generate charts.
ConnectorsAdd sources, run indexing or CRM sync, open memory and evaluation.
ChartsBrowse saved charts and review suggested charts.
DashboardsCreate dashboards, add saved charts, and manage suggested dashboards.
PreferencesUpdate theme, language, and Slack settings.
Connector-specific Memory and Evaluation pages open from the ⋯ menu on each connector row.

​
See the main workspace

This view shows the main app shell, including the left navigation and the suggested question state many users see when starting a new chat. Vizkraft workspace with the sidebar and suggested questions visible

​
Typical workflow

  1. Add a connector in the Connectors area.
  2. Run indexing or HubSpot catalog sync so Vizkraft understands your data.
  3. Open Chat and pick the connector or connectors you want to use.
  4. Ask a question, then use follow-up suggestions or type the next question yourself.
  5. Rate helpful answers and review any memory updates Vizkraft saves.
  6. Save useful charts to your library.
  7. Add charts to dashboards, apply filters (including AI filters), and save views for recurring use.
  8. Review suggested charts and dashboards when Vizkraft prepares them for you.
  9. Optionally run evaluation on SQL connectors to measure and improve accuracy.

​
What to expect on day one

  • If you open chat before asking anything, Vizkraft can show sample suggested questions in the empty state.
  • After an answer is generated, follow-up suggestions can appear above the message input so you can continue the same thread.
  • Memory notifications can appear when Vizkraft learns something about your connector.
  • Suggested charts and dashboards appear in their respective library pages after a connector has been analyzed.

​
Next steps