01
AI Chat Assistant
Use model assistance with notebook and code context for debugging, explanation, and revision without leaving the workflow.
← The Workbench
AI LAB
Your AI-Powered Data Science
JupyterLab rebuilt around AI-assisted analysis. Ask in plain English, inspect the generated query, and produce a chart inside the environment you choose. Connect approved hosted models or keep inference local when zero egress is required.
Start free on your own machine.
FREE DESKTOP DOWNLOAD
Download Calliope AI Lab for personal, professional, or commercial use without creating a Calliope AI account.
Bring credentials for approved hosted models or connect a local model. Notebook files and data processing stay on the machine; prompts and selected context are sent to the model endpoint you choose.
Go from question to deliverable without leaving the notebook.
AI assistance sits inside the JupyterLab workflow, alongside visible code, queries, results, and the Python ecosystem.
- Ask in context · Generate, explain, debug, and refactor notebook code.
- Notebook generation · Describe an analysis and create a runnable starting point.
- Natural language to SQL · Generate a query while keeping the SQL visible for review.
- Connect approved data sources · Work with PostgreSQL, MySQL, Snowflake, BigQuery, S3, and APIs supported by the environment.
- Build reports · Move from a question to code, charts, and exportable results.
- Python environment · Use pandas, numpy, scipy, scikit-learn, matplotlib, and the bundled package set.
- Reproducible setup · Give team workspaces a consistent baseline.
- Workbench administration · Add SSO, RBAC, central credential handling, and workspace access records for teams.
AI Lab Capabilities
02
One-Shot Notebook Generation
Describe the analysis and generate a runnable notebook starting point from the toolbar, File menu, or keyboard shortcut.
03
Multi-Model Support
Connect approved hosted providers or a local model through Ollama. Bring your own provider credentials and select the endpoint appropriate to the task.
04
Your data, one sidebar away
Connect databases from the sidebar and query them with AI assistance. Register curated datasets for reuse in approved Workbench tools.
05
Purpose-Built Interface
Clean, modern dark theme optimized for long coding sessions. Distraction-free design with streamlined UI that hides unnecessary complexity.
06
Privacy by default
The desktop app keeps notebook execution and files on your machine. Model requests go directly to the hosted or local endpoint you configure. A BYOC Workbench keeps the team environment in your account under the same explicit model-destination rule.
Ask in plain English, get the SQL and the chart.
Connect AI Lab to approved data sources and ask questions. The model generates a query and visualization while keeping the SQL visible for verification.
- “Show me sales trends by region for Q4” · Generate the SQL and visualize the result.
- “Which customers churned last month, and what signals correlate with it?” · Inspect the generated queries and analysis.
- “Generate a correlation matrix for these variables” · Create reviewable pandas code and a visualization.
Why teams choose it.
Calliope AI Lab gives teams a familiar Jupyter environment without forcing notebook execution into a third-party notebook service. The organization chooses where the environment runs and which model endpoints it may contact.
Your environment, explicit destinations
Run on a desktop or in your own cloud, connect the data sources you approve, and keep model traffic local when the workload requires zero egress.
The full stack, no setup
JupyterLab 4, 300+ packages bundled, plus the 400,000+ on PyPI. Natural-language SQL and one-shot notebook generation built in.
Desktop or team Workbench
Start free on desktop with your own model access. Move to Workbench for SSO, RBAC, central credential handling, and workspace access records.
