LoginSign Up
← Back to all sparks
L

Lightdash

ANALYTICS
Velocity5.0

Lightdash goes mobile, with Ask AI, custom charts and home-screen widgets on iOS and Android.

business intelligencesemantic layerai agentsmobilecustom visualization
◆Current state
Lightdash is an open-source BI layer turning its semantic layer into a workspace for AI agents. Its new iOS and Android apps show dashboards, saved charts, custom chart types and data apps, let users ask the AI agent questions, work offline and pin charts as home-screen widgets. October made Chart Studio available to all cloud users and added reusable agent skills.
◆Where it's heading
The AI agent is becoming the main way into Lightdash: first through skills and MCP, now through a phone app where Ask AI sits beside dashboards. Each surface reuses the same semantic layer and custom chart types rather than building separate mobile views.
◆Prediction
Expect mobile to pick up agent features such as skills or scheduled deliveries, and more ways for agents outside the app to reach the semantic layer.

◆Recent moves

  1. 3d ago

    📱 Lightdash Mobile

    Native iOS and Android apps bring dashboards, custom charts, data apps and Ask AI to phones, with offline use and home-screen widgets. It extends the agent-first direction of skills and Chart Studio to a new surface rather than changing what Lightdash does.

    View source ↗
  2. 23d ago

    🧩 Chart Studio, now available for all cloud users

    Chart Studio, which turns a described chart into a reusable chart type, is now available to all Lightdash Cloud users and can be driven from Claude through the MCP. It takes September's custom-chart spark from preview to general availability.

    View source ↗
  3. 1mo ago

    🎓 Teach your AI agents reusable skills

    ⚡ SPARK

    Skills let teams write an analysis procedure once and have the agent run it via a slash command, with arguments. It is the clearest step yet in Lightdash turning its agent from a chat feature into a configurable operator, following the Linear/Jira issue export.

    View source ↗
  4. 1mo ago

    🪆 Explore nested and repeated columns

    BigQuery and Databricks users can explore nested and repeated columns directly, with array fields queried per element and warnings for metric inflation. It removes a modeling chore that previously forced flattening upstream.

    View source ↗
  5. 1mo ago

    ✅ Promote, own, and find verified content

    Verified content gains a review queue for promoting personal work, duplicate detection on submission, dashboard owners, and verified filters in search. This governance layer decides what the growing agent and self-serve surface should treat as trusted.

    View source ↗
  6. 2mo ago

    🔌 Test your warehouse connection before you save it

    A Test connection button checks warehouse credentials without recompiling the project. A small setup convenience.

    View source ↗