Skip to main content
Sourcetable’s AI can analyze your data to surface insights that would take hours to find manually. It examines your spreadsheet and identifies trends, relationships, anomalies, and opportunities.

​
Running an analysis

Open the AI chat and select Analyze or EDA (exploratory data analysis) mode, or just ask a question:
  • “Analyze this dataset and tell me the key findings”
  • “What patterns do you see in this sales data?”
  • “Run an exploratory data analysis on this dataset”
  • “Are there any outliers in column D?”

​
What the AI examines

When you ask for an analysis, the AI may look at:
  • Descriptive statistics — mean, median, standard deviation, min, max, quartiles
  • Distributions — how values are spread across your data
  • Correlations — relationships between columns
  • Trends — changes over time
  • Outliers — unusual values that don’t fit the pattern
  • Missing data — gaps and their potential impact
  • Segmentation — natural groupings in your data

​
Data science tools

Under the hood, the AI uses Python data science libraries to perform analysis:
  • pandas — data manipulation and aggregation
  • NumPy — numerical computations
  • SciPy — statistical tests and scientific computing
  • scikit-learn — machine learning models and clustering
  • StatsModels — statistical modeling and hypothesis testing
You don’t need to write any code — the AI handles it. But if you want to see or modify the Python code, the AI can show you what it ran.

​
Machine learning

Sourcetable includes TabPFN, a machine learning model that can make predictions and classifications directly in your spreadsheet:
  • Predict mode — forecast numerical values based on patterns in your data
  • Classify mode — categorize rows based on features in your data
Select the Predict or Classify mode from the mode dropdown and describe what you want to predict or classify.

​
Example use cases

  • Revenue forecasting based on historical trends
  • Customer segmentation by purchasing behavior
  • Anomaly detection in financial transactions
  • A/B test analysis for marketing campaigns
  • Churn prediction based on user activity data
  • Market basket analysis for product recommendations