What Data & Graph does
When enabled, your Agents can perform four types of computational work:- Complex calculations: Financial models, statistical analysis, scenario comparisons, unit conversions. Anything that benefits from precise computation rather than estimation.
- Data analysis: Upload a spreadsheet, CSV, or data file, and your Agent reads, processes, and analyzes it. Pivot tables, trend analysis, outlier detection, and data cleaning.
- Charts and visualizations: Bar charts, line graphs, scatter plots, heatmaps, and more. Your Agent generates visualizations directly in the conversation and iterates on them based on your feedback.
- Data transformation: Clean messy data, merge datasets, reformat files, extract patterns, and prepare data for further analysis.
How it works
Toggle Data & Graph on from the chat input bar (the Python logo icon). A blue pill appears confirming it is active. Your Agent runs code in a secure Python environment with common data science libraries pre-installed (pandas, numpy, matplotlib, seaborn, and others). When your request requires computation, your Agent writes the code, executes it, and presents the results. You see the output (calculations, tables, charts) directly in the conversation. Describe adjustments and your Agent iterates on the spot.When to use Data & Graph
Data & Graph is valuable whenever your work involves:- Financial analysis: Variance calculations, ROI projections, budget comparisons.
- Data exploration: Understanding patterns in uploaded datasets or exported reports.
- Report preparation: Creating charts and tables for presentations or documents.
- Quality checks: Verifying calculations, cross-referencing numbers, validating data.
Vaults
Store data files for repeated analysis
Chat
Where Data & Graph comes to life