FLOOKUP AI DOCUMENTATION

Introduction to Flookup AI

Enhance data cleaning in Google Sheets with Flookup AI, an advanced solution for automated data quality. Flookup AI processes and refines messy data with minimal input. Rather than relying on traditional fuzzy matching algorithms, Flookup AI responds to natural language prompts to carry out advanced data cleaning and validation, based on your instructions.
Whether you are fixing typos, harmonising categories or preparing sheets for analysis, this AI data cleaning tool adapts to your intent and streamlines the process.

Flookup AI automatically selects the optimal AI model based on your data complexity, ensuring cost effective processing while maintaining high quality results. Simple datasets use efficient models, while complex data gets the power of advanced reasoning models.

Data Cleaning with Flookup AI

Go to Extensions > Flookup Data Wrangler > Intelligent data cleaning in your spreadsheet menu.
Select the data cleaning mode you wish to run.
Enter your own OpenAI™ API key. You may store or erase the key securely within your Google account.
Highlight a range of data to analyse and click "Grab selected range".
Enter your prompt, referencing columns by number.
Click an empty cell to indicate where results should be displayed.
Click Submit data cleaning prompt.

Data Cleaning Modes

Fuzzy match: Compare data from two columns and return best matches.
Remove duplicates: Remove duplicates and return only unique values.
Standardize data: Adjust case, trim spaces, correct misspellings and ensure numeric consistency.
Transform data formats: Modify date formats, convert measurement units and apply user specified changes.
Fill in missing data: Fill missing data with placeholders or computed values.
Remove common outliers: Remove outliers and return the cleaned dataset.

Cost Transparency and Processing Intelligence

Intelligent Model Selection

Flookup AI automatically analyzes your data and selects the most cost effective model:

Simple datasets: Uses efficient models for basic cleaning tasks.
Complex datasets: Automatically switches to reasoning models for datasets with mixed data types, wide tables or unique values requiring deeper analysis.
Large datasets: Uses advanced models for datasets over 50,000 tokens to ensure quality.

Real Time Usage Tracking

After each processing session, you will receive detailed information including:

Token usage: Exact count of input and output tokens consumed.
Cost breakdown: Precise cost calculation including cached token savings.
Model used: Which AI model was selected for your specific task.
Processing summary: Number of chunks processed and completion status.

Intelligent Data Processing

Dynamic chunking: Automatically splits large datasets into optimal chunks based on token limits.
Context preservation: Maintains conversation history for consistent results across chunks.
Error recovery: Automatic retry logic with exponential backoff for failed requests.
Data validation: Comprehensive checks ensure your data stays within sheet boundaries.

Flookup AI Usage Policy

Flookup AI is designed to deliver efficient, cost-effective and high-quality data cleaning by leveraging advanced model selection and advanced data chunking. The system automatically chooses the optimal OpenAI model for your data's complexity, ensuring you get the best results for your processing budget.

Cost Transparency: Real-time tracking shows your token usage and costs after each session, so you always know your remaining capacity and spending.
Advanced Processing: Large datasets are automatically split into manageable chunks, with context preserved for accuracy and consistency.
Maximized Value: Advanced processing features help you clean as many rows as possible within your monthly quota, based on your OpenAI API usage.

To get the best results from Flookup AI, follow these prompt-writing guidelines:

Be specific: Clearly describe the task.
One task at a time: For complex jobs, break them into separate prompts for better accuracy.
Use data cleaning keywords: Short, structured instructions are more effective than long explanations.
Address edge cases: Specify if certain values or headers should be excluded or handled differently.
Iterate as needed: If results are not perfect, refine your prompt and try again.

For more tips and best practices, see our comprehensive AI data cleaning guide.

Flookup AI Privacy Statement

Your privacy is important to us. We collect personal details solely for subscription management and service improvement. All data processed by Flookup AI is handled with strict confidentiality, transmitted securely and is not stored externally. We implement robust security measures to protect your information. You retain full rights over your data, including the right to request deletion. For more details, please see our full Privacy Policy.

For any concerns or requests regarding your data, please contact us.