The record stays whole. The single node does not.
DataVeritas makes quality records tamper-evident: test protocols, batch releases and audit trails are hashed at the source, sealed by named authorities and spread as shards across many nodes. No operator has to store everything — and no one can change anything unnoticed.
Quality management runs on evidence. Evidence runs on trust.
Every batch, every calibration, every deviation leaves a record behind. In the audit, one question decides: can you prove this record has not changed since it was created?
Today: central archives
- Whoever holds admin rights on the QMS can change entries after the fact — including the audit trail itself.
- Proof of integrity is a claim made by the archive operator, not a fact anyone can check.
- Suppliers, labs and customers keep separate copies; reconciliation happens by e-mail and PDF.
- Retention periods of ten years and more outlive systems, vendors and database schemas.
With DataVeritas: verifiable immutability
- The hash is created where the measurement happens — signed with the key of the responsible party.
- Named authorities seal blocks; every signature is attributable to a legal identity.
- Data lives as shards on many independent nodes; a tampered shard stands out immediately.
- Auditors check for themselves: fetch shards, rebuild the record, compare the hash with the chain.
Not “trust our archive”, but “check it yourself — in milliseconds.”
Every QM team knows the principles. DataVeritas makes them technically verifiable.
ALCOA+ is the established yardstick for data integrity in regulated environments — from GMP and medical devices to automotive supply. Here is how the architecture maps to its nine principles:
Attributable
Every entry carries the submitter’s signature, every block the sealing authority’s.
Legible
The original stays readable in its source system; the chain secures its fingerprint, not its format.
Contemporaneous
Hash and signature are created at the moment of measurement; block time pins it down every 2.4 s.
Original
Rebuilt shards must match the sealed hash bit for bit — otherwise it is not the original.
Accurate
Diverging copies are detected, isolated and replaced from clean ones; the attempt stays in the log.
Complete
The hash chain exposes gaps: a removed block breaks every link that follows it.
Consistent
All parties see the same chain in the same order — there is no second version of the truth.
Enduring
Per-shard redundancy instead of one archive; failures are re-replicated automatically.
Available
Any authorised auditor can fetch from any nodes at any time and verify independently.
Note: DataVeritas is a technical concept that supports these principles. Compliance with specific standards or regulations (e.g. ISO 9001, EU GMP Annex 11, 21 CFR Part 11) only follows from validation in the respective context of use.
Five steps. None of them asks for blind trust.
The document stays where it is today. DataVeritas makes sure it never changes unnoticed.
Hash at the source
A test protocol, batch release or handover is hashed where it is created and signed with the submitter’s key.
Authority signs the block
The validator on duty orders the entry into a block and signs the header, which carries the hash of the block before it.
Split into fragments
The block body is cut into shards. Each shard gets its own hash; all shard hashes roll up into the header’s Merkle root.
Spread across nodes
Every shard lands on several nodes, picked so that no single operator holds all copies of one fragment.
Rebuild and re-check
An auditor pulls shards from any nodes, rebuilds the record and compares the hashes against the signed chain.
Proof of Authority meets sharding. Attributable and scalable.
A fixed set of named authorities — labs, plants, notified bodies, regulators — takes turns sealing blocks. Storage works like BitTorrent: each node keeps a part, the network as a whole keeps everything.
The whole chain, independent of network size. Audit anchor and seed for new shard distributions.
Holds a share that shrinks with every node that joins, and validates in full through peer requests.
A scaled node with a hard storage cap — a plant PC or an edge box on the line.
Headers and inclusion proofs only. Verifies documents, stores no shards.
Wherever a record has to outlive the system.
Test protocols and batch releases
Measurements, test results and release decisions are signed at the moment of measurement. A batch release can be traced bit for bit years later.
Who changed what, when
Changes to master data, recipes or specifications become entries of their own on the chain — verifiable years later without trusting the archive.
Handovers between partners
Goods receipt, transport, transfer, incoming inspection: every handover between supplier, carrier and inspection body becomes shared, undisputable evidence.
Evidence for regulators and certifiers
Regulatory evidence that no single party can quietly rewrite. Authorities and notified bodies can check along as light nodes.
News and background
Sharding in numbers — every new node lightens the load for everyone
At 24 nodes each one holds about 61 % of the chain, at 1000 nodes only 3 %. We walk through the DataVeritas storage model step by step.
Why quality records need a chain
Quality management stands or falls with one question — has this record changed since it was created? DataVeritas turns that claim into a verifiable fact.
Quality data that speaks for itself.
You are responsible for quality, compliance or a supply chain with many partners? Let’s talk about how DataVeritas fits into your existing system landscape.