We Built the Data Tools We Wished Existed
Refinata started from a simple frustration: every CSV and PDF tool on the internet asks you to upload your files to their servers. We thought that was wrong — especially for sensitive data. So we built tools that run entirely in your browser.
Why We Built Refinata
Three principles guide every decision.
Privacy Is Not a Feature. It's the Architecture.
Most "free" online tools upload your files to a server, process them, and (sometimes) delete them. We thought that was backward — especially for financial data, HR files, and legal documents. So we built Refinata on a simple constraint: files can never leave your device. Every tool runs in your browser's memory via WebAssembly.
Speed Without a Server
DuckDB WASM can process 1 million rows in under 2 seconds in a browser tab. PyOdide boots a full Python environment in < 4 seconds. This is the promise of WebAssembly: server-class performance, client-side. Files never travel over the network, so there's no upload bottleneck, no server queue, no rate limits.
Free by Default, Pro for Power
All 100+ tools are free, forever. Refinata Pro ($9/mo) unlocks AI-powered features: natural language queries via Gemini 2.5, smart auto-clean, and AI summaries. Even in Pro, only column names and statistics (never row data) are sent to the AI model — your actual data rows stay private.
What Powers Refinata
We don't build custom engines. We use the best open-source WebAssembly libraries and compile them to run in your browser.
DuckDB WASM
In-browser SQL engine
Handles massive CSV files without upload. Powers all SQL queries, GROUP BY, JOINs, and window functions — zero backend.
PyOdide
Python 3 in the browser
Full CPython + Pandas + NumPy compiled to WASM. Runs real Python data science workflows offline, with no install.
PDF.js + pdf-lib
Local PDF engine
Mozilla's PDF.js for rendering and pdf-lib for creation. Merge, split, compress, and redact documents without a server.
Tesseract.js
WASM OCR engine
Tesseract compiled to WebAssembly for in-browser text extraction from scanned PDFs and images — no cloud OCR API.
Our Privacy Commitment
Refinata processes every file locally in your browser's memory using WebAssembly. We have no server that receives your file contents. When you close the tab, the data is gone. This is not a policy we can breach — it's the architecture.
File Upload to Server
Never
File Storage
None — memory only
Data Sent to AI
Column names & stats only