Getting Started
What Analytics Agent is, first sign-in, the interface, your first report.
Reports & Documents
Documents built from blocks, the 24-column grid, block help, filters and click-to-drill, PDF export, JSON files and real-time updates.
Block Types
38 content types: charts, tables, metrics, text and Jupyter Notebook.
Charts
28 chart types: from lines and bars to funnels, maps and relationship diagrams — with reference lines, intervals and value labels.
Styling
13 override types: colors, formats, links, styles.
Data Sources
SQLite, PostgreSQL, ClickHouse and your own Python scripts: connecting, schema, result caching; SQLite databases from the Lakehouse are found automatically.
Queries
SQL queries: execution, validation, result caching, performance.
Variables & Templates
Scriban templates, document variables, date selectors.
Agent Integration
Chats and the sidebar on every page, tasks in plain language, the prompts queue.
Validation & Auto-Fix
Validation of documents, blocks and queries; background validation.
Jupyter Notebooks
Python in reports: papermill, pandas, Matplotlib, templates.
Lakehouse
Parquet, SQLite, Excel, PDF, Markdown and logs: browsing and queries.
Analysis Templates
A library of 257 ready-made solutions in two groups — Business and Finance: forecasts, segments, scoring, anomalies — plus a 50-lesson analytics course and demo data for every template.
Installation
Docker Compose, persistent volumes and environment variables, first sign-in.
Alerts
Query rules, 24 notification channels, silences and history.
Python Scripts
Extending with your own scripts: your own data source, result processing, an alert condition, a delivery channel.