Data Tools
Local-first ML tools - cluster logs, explore CSVs, generate patterns, simulate A/B tests, and analyze performance profiles entirely in your browser.
Upload a CSV, Parquet, or JSONL file and run SQL queries on it — powered by DuckDB WASM. No server. Your data never leaves your browser.
Paste raw log lines and instantly cluster repeated patterns using TF-IDF and DBSCAN - runs 100% in your browser via WebAssembly.
Upload a CSV to auto-detect column types, spot missing values, and fit a linear regression - all locally in your browser.
Provide match and non-match string examples; a decision tree learns the pattern and exports it as Rust, JavaScript, or regex.
Train a logistic regression model on A/B experiment observations and use what-if sliders to predict conversion probability.
Upload a Chrome DevTools trace JSON to run PCA and identify which event categories contribute most to latency.



