Basedash

The core of Basedash is a natural language BI interface that turns a written question into a chart. You describe what you want to see, and the platform chooses the chart type and formatting, writes the SQL, and shows it alongside the result. Follow-up questions adjust the same chart, which keeps a line of inquiry moving without rebuilding a query each time.

Dashboards collect charts into a shared view your team can open and scan. Insights push further by having agents watch your metrics around the clock, flag what moved, and explain why in a short brief. You schedule recurring reports, or trigger one the moment your data changes, so you monitor the business much like you monitor production.

Basedash lets you define metrics, segments, and business logic once as reusable models, so "active customer" means the same thing in every chart and every chat. You can pull in existing dbt projects rather than rebuild them, and add team context that the AI remembers from conversation to conversation. One metric definition then feeds dashboards, chat, and embedded views alike.

Connect a source and Basedash provisions a managed DuckDB warehouse that keeps every table synced and ready for agents to query, with storage that scales and point-in-time recovery. It pulls in Salesforce, HubSpot, Stripe, Shopify, and hundreds more without you building a pipeline, and it connects to Snowflake, BigQuery, or Databricks if those are already in place.

You can embed dashboards and AI chat inside your own product using typed React components and server-side JWT helpers. White-label the view to your brand and scope each customer to their own rows through JWT single sign-on, which gives end users self-serve analytics without pulling engineers off the roadmap. That turns reporting into a product feature rather than an internal tool.

Basedash exposes an MCP server so you can connect Claude, ChatGPT, or Cursor and ask questions from the tools you already work in, and it ships a Slack app that answers and posts automations in a channel. A REST API covers programmatic access for your own apps. All three paths run on the same models and permissions, so answers stay governed no matter where they're asked.

Every question runs under your permissions, lands in an audit log, and shows the data behind its answer, which addresses the gap left when staff query their own AI tools. The platform adds approval for AI-run write queries, SSO with SAML and OIDC, SCIM provisioning, and row-level security. It holds SOC 2 Type II, offers self-hosting in your own VPC, and never trains models on customer data.