# Dataweaver

> Dataweaver is a local-first AI data analyst desktop app that analyzes spreadsheets, CSVs, SQL databases, and cloud logs on your machine, with no file uploads and automatic PII redaction. Closed-source, published by TechCora.

Dataweaver is a one-stop desktop tool for operations teams and data workers who deal with messy files and data sources daily. Users describe what they want in plain English — "clean up dates", "remove duplicates", "flag missing payments" — and an AI agent handles the rest automatically. No code, no formulas, no training required.

The system is fully agentic with two AI agents: a Transformation Agent for data cleaning, reshaping, and file operations, and an Analyst Agent for generating charts, summaries, and insights. The agent picks the right approach automatically — users just describe the goal.

All data processing runs locally on the user's machine. PII is automatically redacted before AI processing. No data is uploaded to any cloud service. GDPR-compliant by design. Source files are never modified — transforms always create new output files.

Supports bring-your-own-key (BYOK) for AI models and custom LLM integration for enterprise customers.

## Docs
- [Homepage](https://dataweaverapp.com/): Product overview, download, and feature showcase
- [Documentation](https://dataweaverapp.com/docs): Guides and recipes for common data tasks
- [Changelog](https://dataweaverapp.com/changelog): Version history and release notes
- [Pricing](https://dataweaverapp.com/pricing): Free and Pro plan details
- [Contact](https://dataweaverapp.com/contact): Support and sales inquiries
- [Privacy Policy](https://dataweaverapp.com/privacy): Data handling and privacy practices
- [Terms of Service](https://dataweaverapp.com/terms): Legal terms

## Data Sources
- **Files**: CSV, Excel (.xlsx), JSON, Apache Arrow — drag and drop, handles millions of rows
- **Databases**: Connect to any SQL database, browse tables, run queries through the AI agent
- **Cloud Logs**: GCP Cloud Logging — search, analyze, and chart log data with natural language
- **REST APIs**: Connect to HTTP services, the agent discovers endpoints and schemas automatically

## Key Features
- Plain English data transformation — describe what you want, no code required
- Fully agentic AI with Transformation Agent and Analyst Agent (generates interactive charts)
- Three agent modes: Quick (simple cleanup), Standard (day-to-day), Expert (complex files)
- File import with multi-sheet Excel support and automatic format detection
- Database connections with schema search and read-only SQL queries
- Cloud log exploration with severity filtering, time-range analysis, and pattern detection
- HTTP service connectors with automatic endpoint discovery
- Web search integration — agent can look up information while working on your data
- Automatic PII redaction before AI processing
- GDPR-compliant by design — all processing runs locally
- Source files never modified — immutable originals, new output files always
- Workflow save and replay — run the same cleanup on next month's file
- Command palette for quick access to files and actions
- Bring your own AI model key (BYOK)
- Custom LLM integration available for enterprise

## Use Cases
- **Data Cleaning**: Fix inconsistent formats, remove duplicates, standardize dates in CSV and Excel files
- **Database Queries**: Ask questions about your database in plain English, get charts and tables
- **Cloud Log Analysis**: Search GCP logs, investigate incidents, chart error patterns over time
- **Data Validation**: Flag missing values, validate ranges, check data quality
- **Recurring Workflows**: Save cleaning workflows and replay on monthly/weekly data files
- **PII-Safe Analysis**: Automatically redact sensitive data before AI processing
- **Cross-Source Analysis**: Work with files, databases, and logs in the same conversation

## Comparison
- **vs ChatGPT/Claude for data**: Dataweaver works directly on local files and databases, no copy-paste, handles million-row files, preserves privacy with PII redaction, generates interactive charts
- **vs Excel**: Instant processing of large files, AI-powered transforms instead of manual formulas, immutable originals (no accidental data loss)
- **vs Python/pandas**: Plain English interface, no environment setup, automatic error recovery, built-in charts
- **vs cloud ETL tools**: No data uploads, instant setup, desktop-native, free tier available

## Company
- **Developer**: TechCora (https://techcora.in)
- **Founder**: Abin Paul
- **Product Lead**: Gagan Raj Bojja
- **Website**: https://dataweaverapp.com
- **Contact**: https://dataweaverapp.com/contact

## Optional
- [GitHub](https://github.com/techcoracorp): Organization page
- [Twitter/X](https://twitter.com/techcoracorp): Product updates
- [LinkedIn](https://linkedin.com/company/techcoracorp): Company page