OPEN-SOURCE BUSINESS INTELLIGENCE

Metrics your whole team can build on.

Define your business metrics once. Build dashboards and ask questions through an AI assistant, using the same metric definitions and access rules.

LeapView Visual Showcase dashboard with KPI cards, revenue chart, order status donut chart, delivery bars, product categories, and an orders table

Your business intelligence should belong to your business.

Your metrics and dashboards capture how your business works. You should be able to understand that logic, build on it, and choose where it runs.

We’re building LeapView so the full BI product—including governance, security, and AI integrations—remains open source. Analytics are defined in code, with open interfaces and the option to host it yourself.

Our goal is to keep you free to choose where it runs, who operates it, and which analytics tools you use as your business evolves.

We want you to stay because LeapView keeps being useful.

Contribute on GitHub

Connect the data
you already work with.

Start with PostgreSQL, Parquet files in S3, or an Excel spreadsheet. Use that data to define your metrics and build dashboards.

Connect your data

Analytics as code. Changes you can review.

Keep connections, transformations, metrics, and dashboards in version control. Review changes with your team and reuse metric definitions across dashboards.

01 / CONNECTConnect to your data.

sales-project / connections / olist.yamlYAML
Loading olist.yaml…

Share your analytics. Keep control of access.

Use your company’s existing sign-in, decide who can query which data, and record security changes and query activity.

Sign in with your company account.

Connect your identity provider with OIDC single sign-on. Keep users and groups in sync with SCIM.

Explore single sign-on

Decide who can see what.

Set roles and permissions, filter rows, and mask sensitive columns. Apply access policies to queries from dashboards, APIs, and AI agents.

Explore access controls

Trace access changes and query activity.

Review security and administrative changes. See which user or service ran a governed query and which resource it used.

Explore audit logs

How LeapView works.

Dashboards, AI agents, and APIs query a common set of metrics and dimensions defined in semantic models. DuckDB runs the queries. PostgreSQL stores application state, while DuckLake manages analytical snapshots in Parquet.

DASHBOARDSVISUALS AGENTSMCP
SEMANTICMODELS
DUCKDBCOMPUTE
POSTGRESSTATE DUCKLAKEPARQUET

INTERFACES

Dashboards

Explore metrics through interactive charts and tables.

INTERFACES

Agents and APIs

Query business data using the same metric definitions.

Agents MCP API CLI

SERVING

Semantic models

Shared metrics and dimensions with access policies.

COMPUTE

DuckDB

Execute analytical queries and data refreshes.

STORAGE & STATE

PostgreSQL

Store application state, project records, and access configuration.

STORAGE & STATE

DuckLake

Manage analytical snapshots in Parquet, with catalog metadata in PostgreSQL.

Your infrastructure. Your definitions. Your choice.

Choose where LeapView runs, which AI provider it uses, and how you share metric definitions with other tools.

Run it on your infrastructure.

Host LeapView yourself using the public Docker image and Compose package. The software is available under Apache License 2.0.

Explore self-hosting

Choose your AI provider.

Connect built-in chat to an OpenAI-compatible endpoint, or use an external AI assistant through MCP. Both access analytics through LeapView’s tools and access policies.

Explore AI integrations

Keep your definitions in code.

Maintain readable, version-controlled analytics definitions. Import and export semantic models with Apache Ossie, and connect other tools through documented APIs and MCP.

Explore model import and export