INTELLIGENCE
Guides
How to Get Started with Polars in Python
Polars is a Python DataFrame library written in Rust that processes data 3 to 14 times faster than pandas depending on the operation. It uses Apache Arrow's columnar memory format and a lazy evaluation engine that optimizes queries before running them. Install it with one pip command; Polars reads CSV, Parquet, and JSON files directly, and covers filtering, groupby, joins, and aggregation with the same analytical workflow pandas users already know.
How to Analyze CSV Files with DuckDB
DuckDB is a free, zero-configuration SQL database that reads CSV files directly without any import step. Download the single CLI binary, run duckdb from your terminal, and query any CSV file with standard SQL immediately. No Python, no database server, and no schema definition are required. DuckDB detects column types automatically, supports aggregations and joins across multiple files, and handles datasets from a few rows to several gigabytes on a standard laptop.
How to Set Up Metabase Metabot AI
Metabase Metabot AI is an AI assistant built into the Metabase analytics platform that lets users query databases in plain English without writing SQL. Setting it up requires connecting a database, enabling Metabot with an Anthropic API key, and adding table and column descriptions to improve answer quality. The full process takes under 30 minutes for both self-hosted and Metabase Cloud deployments.
How to Analyze Data with Claude AI
Claude AI can analyze CSV and Excel files directly from the chat interface without any setup, database configuration, or coding. Upload a file, describe what you need in plain English, and Claude returns calculations, summaries, and charts in seconds. Its 200,000-token context window handles roughly 150,000 rows of business data in a single session, making it one of the few AI tools where a complete year of sales or operations data fits in one conversation without truncation.
How to Use Gemini AI in Google Sheets
Gemini AI in Google Sheets gives analysts three distinct tools: a conversational sidebar for ad hoc dataset questions, an =AI() cell formula for row-level classification and extraction, and Connected Sheets for natural-language queries against BigQuery at scale. Each feature requires a Google Workspace Business Standard plan or higher. The sidebar handles trend spotting and chart generation; the =AI() function processes thousands of rows without writing formulas manually; Connected Sheets queries billions of rows from BigQuery without importing data.
How to Set Up Microsoft Fabric for Data Analysis
Microsoft Fabric is a unified analytics platform that combines data ingestion, storage, transformation, and reporting into a single workspace. To set it up for data analysis, start a 60-day free trial at app.fabric.microsoft.com using a work or school Microsoft account, create a workspace assigned to trial capacity, build a lakehouse to store your data, and connect Power BI for reporting. AI features like Copilot and Data Agent require a paid F64 or higher capacity.
How to Set Up ChatGPT for Excel
ChatGPT for Excel is a native Office Add-in launched in March 2026 that lets you build financial models, run scenario analysis, and explain formulas using plain-language prompts. It requires ChatGPT Plus ($20/month) or higher and an active Microsoft 365 subscription. Available in the US, Canada, and Australia, setup takes under five minutes through the Excel Add-ins store, though performance on complex models varies.
How to Get Started with Golden Analytics
Golden Analytics is a new AI-native business intelligence platform that launched in April 2026 with $7 million in seed funding. Built by former Tableau CPO Francois Ajenstat, it lets analysts go from a raw CSV to a shareable dashboard without manual configuration. The platform's defining feature, the Slider of Autonomy, lets users choose how much work the AI does. Early access sign-up is open at goldenanalytics.com.
How to Build Reports with Power BI Copilot
Power BI Copilot lets analysts and business users build complete report pages from plain English prompts, without writing DAX or configuring visuals manually. It works within Power BI service and Desktop, requires a paid Fabric capacity (F2 or higher), and performs best when the underlying semantic model has clean column names and field descriptions. Copilot handles chart selection, layout, and narrative summaries, but accurate output depends heavily on how well the data model is prepared before use.