Data Analyzer

What it does

The Data Analyzer skill is an expert agent designed to process structured and unstructured datasets, extracting meaningful insights and identifying patterns. It performs exploratory data analysis (EDA), statistical testing, correlation analysis, and creates narratives around data findings. The tool transforms raw data into actionable intelligence through rigorous analytical frameworks and data visualization best practices.

When to use it

• Business Analytics: To understand key performance indicators and identify areas for improvement.
• Research Validation: To analyze research data and validate hypotheses.
• Performance Analysis: To assess the effectiveness of strategies or processes by examining relevant datasets.
• Decision Support: To provide data-driven recommendations to inform strategic decisions.

Key capabilities

• Data Profiling: Determines dataset dimensions, column types, missing values, unique values and distributions.
• Data Quality Assessment: Identifies missing data patterns, duplicate records, outliers, inconsistencies, and format mismatches.
• Univariate Analysis: Analyzes the distribution of individual variables, identifying skewness, kurtosis, and outliers.
• Bivariate & Multivariate Analysis: Performs correlation analysis (Pearson, Spearman), scatter plots, cross-tabulations, dimensionality assessment, and cluster tendency analysis.
• Time Series Analysis: Identifies trends, seasonality, cyclical patterns, and anomalies in temporal data.
• Segmentation Analysis: Groups data into segments for comparison across key metrics.

Example prompts

• "Analyze this dataset and provide an EDA report with summary statistics and visualizations."
• "Identify any significant correlations between these variables in the provided data."
• "Perform a time series analysis on this sales data to identify trends and seasonality."

Tips & gotchas

• The skill specializes in structured and unstructured datasets, but performance may vary depending on data quality.
• It’s important to understand that correlation does not equal causation; the agent will document this distinction carefully.
• The tool provides preliminary analyses and insights – further investigation might be needed based on its findings.