Powered by: DuckDB-WASM · WebGPU · Apache Arrow · AES-256-GCM BYOK · SHA-256 Audit Chains
Client-Side Profiling
Column types, distributions, cardinality, null ratios, outliers — computed on upload
BYOK AI Agent & Recipes
Your chosen LLM or 1-click Smart Recipes suggest cleaning steps from metadata only.
WebGPU KNN Imputation
Hardware-accelerated nearest-neighbor imputation for missing values with confidence scores
Entity Resolution
Vector similarity matching: 'Apple Inc' and 'Apple, Inc.' resolved automatically
Visual DAG & Spreadsheet
Drag-and-drop pipeline canvas with 15 node types plus an interactive Data Wrangler grid.
Pipeline Templates
Save any pipeline as a reusable template. Apply to new datasets with a single click.
Synthetic Data Generator
Bootstrap, distribution, correlation-preserving, and differential privacy generation methods
Universal Ingestion
CSV, TSV, JSON, Excel, Parquet, PDF Document Extractor, and cloud buckets (S3/R2/GCS).
Full Cryptographic Audit
Every operation logged with SHA-256 Merkle chain, before/after diffs, and provenance tags
Ready to clean your data?
Free unlimited local compute. Physical zero egress by design.