Ordit creates cryptographically immutable records of automated decisions.
An audit trail is created automatically.
Any modification after ingestion is detectable.
Timestamp
Decision
Type
AI Model
Status
d4e8c21a9...
LOAN_APPR
LightGBM
APPROVED
a7b3f842e...
FRAUD_CK
Isolation Forest
DECLINED
5e9d7c41b...
PRICING
XGBoost
APPROVED
How It Works
Decisions captured automatically
Automated and AI decision systems send records to Ordit via API the moment a decision is made.
Audit trail created
Ordit computes and stores cryptographic receipts and hash chains server-side for each record.
Verify integrity later
Records can be checked for integrity against their original state by any authorized party.
Built for decisions that must be trusted later
Ordit is designed for systems where automated decisions need to remain provable, auditable, and tamper-evident long after they are made.
AI decision auditing & compliance
Scenario
Regulators or auditors examine an automated decision years after it was made and require proof that the outcome is authentic and unaltered.
How Ordit helps
Ordit cryptographically seals each decision output, creating a permanent, verifiable record that can be checked at any time by any authorized party.
- Prove a decision existed at a specific point in time
- Detect any modification after ingestion
- Replace mutable logs with cryptographic evidence
Independent integrity verification
Scenario
A decision record is shared with a third party who needs to confirm that the data presented today exactly matches what was originally recorded.
How Ordit helps
Ordit enables authorized parties to verify a decision record remains byte-for-byte identical to its original state, without relying on internal systems or trust in the producer.
- Verify record integrity without privileged access
- Detect tampering or misrepresentation
- Establish a shared, tamper-evident reference
Version tracking & governance
Scenario
Teams deploy, update, and iterate on models over time and need confidence that historical decisions remain attributable and auditable across versions.
How Ordit helps
Ordit preserves immutable records of decision outputs and deployment metadata, enabling teams to track changes and maintain accountability across versions.
- Attribute decisions to specific model versions
- Audit historical behavior after updates
- Strengthen accountability across deployments
Frequently Asked Questions
Ordit is an immutable decision recording system. It sits alongside automated systems and automatically records decisions as they occur, creating tamper-evident, verifiable records that can be checked months or years later.
Decisions are recorded automatically via API at the moment they are made. There is no manual logging process and no human intervention required. Once recorded, a cryptographic receipt is returned immediately.
Ordit is designed for any automated or semi-automated decision that produces a structured outcome, including:
- AI model outputs
- Rules-engine determinations
- Risk, fraud, or eligibility assessments
- Credit, lending, or pricing decisions
- Compliance or policy enforcement decisions
- Internal operational decisions with audit requirements
No. Ordit does not run models, access inputs, or influence decision logic. Models and rules remain fully owned and operated by the organization. Ordit records the final decision output and related metadata only.
No. Ordit works with any automated decision system, whether it uses machine learning, deterministic rules, heuristics, or a combination of approaches. AI is supported, but not required.
Ordit is used by organizations in regulated, high-stakes, or audit-driven environments—compliance, risk, audit, legal, and engineering teams.
Each decision is cryptographically hashed and linked to the previous record in a hash chain. Any modification breaks the chain and is immediately detectable during automatic verification.
No. Ordit uses hash chains on traditional infrastructure. There is no cryptocurrency, no public ledger, and no distributed consensus.
Only the decision output and related metadata submitted by the organization. Ordit does not store model inputs, internal reasoning, or training data unless explicitly provided.
Organizations control access. Verification can be performed by authorized parties without modifying records.
Ordit is designed to support compliance with data protection and audit requirements. Sensitive data is only stored if explicitly provided, and exports can be configured to exclude personal information.
Integration is a one-time setup. After a decision is produced, a single API call records it automatically. Ordit returns a cryptographic receipt immediately.
Any system capable of making an HTTP request can integrate.
No. Ordit records finalized decisions, not conversations, prompts, or intermediate reasoning steps.
Decision systems continue operating normally. Ordit is a recording layer, not a dependency for decision execution.
The payload is normalized into a consistent JSON representation so hashes and signatures are stable and reproducible.
Each record includes a pointer to the previous record’s chain value, forming a tamper-evident chain.
Organisations can verify their own decisions at any time, and can share a decision hash with auditors/regulators for third-party verification.
Verification confirms integrity and continuity of the record chain. It does not evaluate whether the decision was “correct”.
Yes. Exports are designed for offline verification using the included instructions and signature materials.
The public verifier should avoid PII and return only integrity status. Detailed payload views remain authenticated and protected.