Interactive, real-time dashboards on any data source
Charts, scheduled reports, drilldowns and self-service exploration on SQL, NoSQL, REST APIs, files and documents, queried where they already live. No warehouse to load first. Or describe the dashboard and a Knowi agent builds it.
40+ chart types · AI dashboard generation · cross-source joins without ETL
The dashboard reads the source. Nothing lands in a warehouse first.
Connect the source. Chart it.
Knowi queries nested JSON and arrays in their native shape, so MongoDB and Elasticsearch data is not flattened into a star schema before it can be charted. Sources join at query time.
- SQL, NoSQL, REST APIs, files, object storage and documents
- Cloud9QL joins across sources at query time, no pipeline
- Structured and unstructured data in the same widget
Describe the dashboard. An agent builds it.
Knowi's agents run inside the data layer, against the real schema and the governed business terms. Ask for a dashboard and the orchestrator picks the sources, writes the queries, chooses the chart types and lays out the page. The generated query stays visible and editable.
- Create Dashboard and Create Widget agents handle layout and chart selection
- Natural language queries resolve against governed business terms
- Follow up in the same thread to filter, add a widget or schedule delivery
Live, cached or scheduled. Per dataset.
Queried live at the source, live with results cached, or materialized on a schedule. Streaming sources update as events arrive. Change the choice later without a migration.
- Live query mode with no cached copy in between
- Webhook-triggered and event-driven refresh
- Streaming and log sources including Kafka, Elasticsearch and DynamoDB
- Threshold and anomaly alerts to Slack, Teams, email or webhook
The same dashboard, inside your product.
A dashboard built for your team can be handed to your customers without rebuilding it. One definition, filtered per tenant at query time, with row-level security and SSO. Your users never see Knowi.
- JavaScript SDK or iframe embed, fully white-labeled
- Row-level security and multi-tenant isolation enforced at query time
- SSO and JWT so users never see a second login
- Dataset and dashboard permissions, encryption and access controls
Every chart, on any source.
Time series, geo, pivot and KPI widgets all read the same governed dataset, structured or not. The visual is never constrained by where the data lives.
Movement over time
Line, area, multi-series, moving averages and event overlays for trend and seasonality work.
Shape and spread
Bar, column, stacked, grouped, histogram, box plot and waterfall.
Where it happened
Choropleth, point maps, region maps and custom geoJSON layers.
The detail underneath
Tables, pivot tables, grids and metadata inspectors for drill-down and reconciliation.
The number that matters
Summary cards, gauges, donut charts and scorecards for exec and wall-mounted views.
When the library is not enough
JavaScript widgets, HTML blocks and custom branding for visuals specific to your domain.
Build it by hand, or from one prompt.
Connecting a source, shaping a dataset and laying out the widgets.
70+ sources, joined at query time.
Connect a source in minutes and query it in place. Nothing has to be consolidated into a warehouse before it can be charted.
Questions about dashboards.
Can I build dashboards in Knowi without writing SQL?
Yes. Ask in plain English and the Query and Create Dashboard agents generate the query, pick the chart type and lay out the dashboard. The Visual Query Builder covers the same ground for people who prefer to click rather than type.
What data sources does Knowi support?
SQL databases, NoSQL databases, REST APIs, cloud applications, files and documents, so you can analyse structured, semi-structured and unstructured data in one platform.
Can Knowi join SQL, NoSQL and API data in one dashboard?
Yes. Cloud9QL performs cross-source joins across SQL, NoSQL, APIs, files and documents without ETL, pipelines or data movement, so a single widget can span MongoDB, Postgres and a REST endpoint.
Does Knowi support real-time dashboards?
Yes. Live queries, webhook triggers, streaming sources, incremental updates and event-driven refresh are all supported, including Elasticsearch, DynamoDB and Kafka.
Can I embed Knowi dashboards inside my application?
Yes. Embed via JavaScript SDK or iframe with white-labeling, SSO and JWT authentication, row-level security and custom theming.
What is the best way to analyse NoSQL data?
Use a tool that natively understands nested JSON, arrays and schema-less data. Knowi connects directly to MongoDB, Elasticsearch, DynamoDB and Couchbase, flattens and transforms JSON automatically, and lets you combine that with SQL and API data in the same dashboard.
Which BI tool works with MongoDB natively?
Knowi queries MongoDB directly, including JSON flattening, nested field extraction, array handling and pipeline stages, with no export step and no schema to build first. MongoDB data can be joined with SQL and API data in the same dashboard.
How many chart types does Knowi have?
More than 40, covering time series, comparison and distribution, geospatial, tabular and pivot, KPI and executive widgets, plus custom JavaScript and HTML visualizations.
What is a KPI dashboard?
A KPI dashboard tracks a small set of business metrics on a single screen and updates as the underlying data changes. In Knowi every KPI widget is a live query against the source system, so the number on screen is the number in the database, not a copy loaded overnight.
Can AI build a dashboard for me?
Yes. Describe the dashboard you want in plain English and Knowi's agents write the query, choose the chart type and lay the widgets out. An AI dashboard built this way is an ordinary Knowi dashboard: everything stays editable and the generated query stays visible.
What is a business analytics platform?
A business analytics platform connects to your operational systems, models the data, and turns it into dashboards, reports and alerts that people across the business can use. Knowi does this without a warehouse loading step, querying SQL, NoSQL, REST APIs and documents where they already live.
See it on your data.
Connect a source, join it to another, and put the view on a dashboard. No warehouse project in front of it.
Any datasource · agent-built dashboards · fully embeddable