ProofFrame
Data contracts with verifiable quality evidence.
Scans a dataset against a contract and returns violations, a decision status and a verifiable evidence envelope.
- For
- Data platform and analytics teams
- Engine
- Rust · Python · WASM
- Source code
- Open · Apache-2.0
- Version
- 0.7.2
- Buying
- Written, fixed-price pilot
What is ProofFrame?
ProofFrame is an open-source, Arrow-native data quality tool. It scans a dataset against a data contract and returns a three-valued verdict: accepted, rejected or unknown. The result goes into an Ed25519-signed evidence envelope that can be verified offline without the data. It is written in Rust, with a Python package and a linter that runs in the browser.
From contract to signed evidence, with a three-valued verdict.
Contract compiled against the schema · before the first row
Visualisation. Rows come from the README example; the fingerprint is a SHA-256 of the table shown here, computed in your browser. ProofFrame's own fingerprint protocol differs.
How it works
From the product's own documentation. Each number says where and on which version it was measured.
The contract compiles first
Rules compile against the schema before the first row is read. A rule that does not fit the type is rejected up front.
It scans Arrow batches
Pandas, Polars, PyArrow, CSV, Parquet. Cells never become Python objects; memory and disk limits are stated in advance.
It answers in three values
accepted, rejected, unknown. An unreadable file or exhausted limit never quietly becomes 'accepted'.
The evidence travels with it
Evidence V2 binds fingerprint, schema, contract and plan. Ed25519-signed, verified offline without the data.
- 0.7.2
- Published on PyPI and crates.io. Airflow and Dagster packages ship separately. Source
- 3 states
- Most validators have nowhere to put 'the check did not run'. This one does: unknown. Source
- In the browser
- The engine compiles to WebAssembly. A CSV is scanned in the tab; nothing is uploaded. Source
- 3 operating systems
- CI on Linux, macOS and Windows. Python 3.10–3.13, Rust 1.85. Source
Stack
- Rust
- Apache Arrow
- Python (PyO3)
- WebAssembly
- Ed25519
- Airflow
- Dagster
- GitHub Actions
Try it
pip install proofframe
proofframe suggest orders.csv > contract.json # review, then set "active"
proofframe review orders.csv --contract contract.json --out review-runWhat it does not claim
Not an observability platform. A library and CLI for deterministic checks at an Arrow boundary, with a deliberately narrower rule set than Great Expectations.
Current boundary
The browser linter runs on a local file. Organization-specific integration is a separate pilot.
What a pilot covers
A sample starting scope. Final deliverables and acceptance criteria are fixed in the written proposal.
- Question
- How can a dataset delivery be proven reliable?
- Likely input
- Sample dataset, schema and acceptance rules
- Deliverables
- Data contract draft
- Quality findings and diff report
- Verifiable delivery receipt
Before you decide
Can I buy this product today?
A paid pilot and setup proposal can be prepared. Scope, data access, deliverables, timeline and a fixed price are agreed in writing after a discovery call.
Why is there no price?
Data volume, integration and success criteria differ from job to job. After the call you get a fixed-price written proposal; no paid work starts without your approval.
Do I need to upload live data here?
No. This site has no file upload or accounts. Data access, retention and permissions are set only in the pilot agreement.
What is the first step?
Choose ProofFrame in the form and describe the problem you need to solve. The data and pilot boundaries are then set together.
More in this discipline
- CRM Cleaning APIRust (Axum) · Python
Prepares messy lead lists for CRM import.
- Fuzzy DedupeRust · Python
Near-duplicate detection for names and organizations.
- Scraping PipelinePython · asyncio
Turns public web data into validated output.