← Back to all sparks
O

Omni

ANALYTICS
Velocity5.0

Business intelligence and embedded analytics platform

Omni keeps welding AI into the BI modeling layer, one weekly drop at a time

business-intelligenceai-agentssemantic-modelingembedded-analyticsweekly-cadencegovernance
Current state
Omni ships weekly, and the throughline is AI moving from bolt-on to default: AI-powered visualization annotations and the AI Hub have reached general availability, and external AI context now pulls from Notion. Alongside that runs steady modeling and embedding work — calculation pushdown, dynamic top-N parameterization, approximate aggregates, and timezone overrides in embed URLs.
Where it's heading
The product is converging an analyst-facing BI tool with an agent-assisted one: a Modeling Agent with skills, AI context-management controls, and access-gated AI grants suggest Omni wants AI that respects the semantic model rather than bypassing it. The non-AI work (compute routing, CLI OAuth, publish-draft APIs) is hardening the platform underneath so the AI features have a governed surface to act on.
Prediction
Expect the next releases to keep promoting AI features from preview to GA and to deepen the Modeling Agent's reach into governed datasets; further external-context integrations beyond Notion are the likeliest near-term add.

Recent moves

  1. 3d ago

    AI visualization annotations GA; Notion AI context integration

    Visualization annotations graduate to GA and external AI context now reads from Notion — both fit the arc of making AI a governed, default part of the modeling surface rather than an experiment. Apps being enabled by default for new orgs signals confidence in that bundle.

    View source ↗
  2. 10d ago

    Dashboard spacers/dividers, AI file uploads, Notion integration

    A mix of dashboard layout primitives (spacers, dividers), embedding display controls, and AI file uploads — incremental polish on both the authoring surface and the AI context pipeline that the following week's GA builds on.

    View source ↗
  3. 10d ago

    Calculation pushdown, dynamic top-N modeling, Slack setup

    Modeling-layer gains: calculation pushdown for range references and dynamic top-N parameterization, plus a simpler Slack integration and timezone overrides in embed URLs. This is the platform-hardening that runs in parallel with the AI push.

    View source ↗
  4. 16d ago

    Publish-draft API, approximate aggregates, map viewport config

    A publish-draft-document API endpoint and sketch-based approximate aggregates in routing extend programmatic and performance control over content, consistent with Omni opening more of the platform to automation.

    View source ↗
  5. 1mo ago

    AI Hub and Markdown columns reach GA

    AI Hub reaching GA, with AI skills gated behind required access grants, is a key step in the governed-AI trajectory: the capability ships broadly but only where permissions allow. Markdown columns going GA is the smaller co-release.

    View source ↗
  6. 1mo ago

    Compute routing, calculated-field drilling, CLI OAuth

    Support for compute routing, drilling into calculated fields, and OAuth for the Omni CLI — infrastructure and developer-experience work that broadens how teams query and automate against the model.

    View source ↗