Decision ReceiptGeneric DecisionDR-K7M9-2P4Q
Verified- Decision
- Approved by Risk Committee
- Policy
- Third-party risk — Tier 1, v7
- Evidence
- 3 attestations, 2 documents
- Integrity
sha256:0xdead…beef
How it works
Decisions happen everywhere.
MeshQu sits at the point they are made. The Decision Assurance architecture in three moments.
Three moments
The proof is captured while the decision is still live.
- A transaction is evaluatedPre-flight policy check
- A model produces a resultThreshold + override logic
- An approval is requestedRouting + actor binding
Before execution
- Inputs capturedExact values, hashed
- Policy boundVersion + clause applied
- Outcome sealedSigned at the moment
At execution
After the fact is audit. At execution is control.
Anatomy of a decision
Every decision has structure.
Four fields, captured together. ACTOR, POLICY, CONTEXT, EVIDENCE — bound at the moment the call is made.
ACTORThe system or person making the decision
POLICYThe rules or logic applied
CONTEXTThe situation at the time
EVIDENCEThe data used
The receipt
What comes out is a Decision Receipt.
Short. Structured. Unambiguous. The four inputs, the outcome, the actor, and a signature.
- SignedBound to a published key. Verifiable offline.
- VerifiableNo API, no account, no infrastructure trust required.
- ReplayableSame inputs, same policy — same result.
- Captured laterLogs, dashboards, memory
- ReconstructedAfter the fact
- Dependent on systemsTrust the producer
Without MeshQu.
- Captured at executionInputs + policy + outcome
- SealedSigned at the moment
- ReplayableAnywhere, any time
With MeshQu.
Decisions form chains
A receipt proves a decision. A chain proves a workflow.
Steps are hash-linked as they happen — screening, review, disposition. When the workflow closes, the chain is sealed.
4 receipts1 provable outcome
- Step 01Intaker-001policyokresultvalidseal◎
- Step 02Classifyr-002policyokresultvalidseal◎
- Step 03Policyr-003policyokresultvalidseal◎
- Step 04Outcomer-004policyokresultvalidseal◎
Trust posture
Verifiable without trusting MeshQu.
Receipts are signed against published keys. Anyone can verify a receipt offline — no API, no account, no infrastructure trust required. Receipts can be anchored to a public transparency log; the proof outlives us.
Questions
Common objections
Is MeshQu just an audit log?
No. Logs record events after they happen. MeshQu captures the decision at execution time, including the policy, context, evidence and outcome.
Do we need to replace our existing systems?
No. MeshQu sits alongside existing systems and captures the decisions they produce.
Can MeshQu work with human decisions?
Yes. Human approvals, automated rules and AI-assisted decisions can all produce the same Decision Receipt.
Can decisions be verified without trusting MeshQu?
Yes. Receipts can be independently verified, including against public keys and external anchoring where configured.
What does replayable mean?
It means the original decision can be reproduced against the same inputs, policy version and context.