Machine Commerce Attestation · Open Standard

Agents spend. IONDRA proves it.

Every agent payment becomes a receipt anyone can verify: what was requested, what money moved, what was delivered. No account. No need to trust us.

Observability records what a platform saw. A receipt is what a stranger can check.

Observability

Recording

  • Lives in the vendor's store
  • Written in the vendor's format
  • You trust the vendor

Attestation

Receipt

  • Hash-only record in a public log
  • Ed25519-signed Merkle checkpoint
  • One-command verify
  • No account

What a receipt binds

Request

What the agent asked for.

request_hash · SHA-256

Payment

What money moved, on which rail.

payment_hash · SHA-256

Delivery

What came back.

delivery_hash · SHA-256

Hashes only. Bodies and PII are never stored. Mandate binding is next.

How it works

01

Hash at the edge

Request, payment and delivery are hashed where they happen. Only digests leave.

02

Append to the public log

Each hash-only record is appended to a public, append-only log.

03

Seal in a signed Merkle checkpoint, witnesses cosign

Records are sealed under an Ed25519-signed Merkle root; witness keys cosign the checkpoint.

04

Anyone verifies with one command

Recompute the hashes, check the inclusion proof and the signatures. No account.

Verify it yourself

git clone https://github.com/1omega/mca-verify

Walkthrough: /proof/0

Today: two witness keys, both operated by IONDRA. An independent third-party witness is the next milestone. Every checkpoint lists who signed it.

The IONDRA agentic suite

Enterprise Preview

ACA-X — Agent Commerce Governance

Buy-side governance for the agent economy. As AI agents become paying customers of the internet, ACA-X gives organizations visibility into everything their agents call, consume, and are asked to pay for — recorded to the same tamper-evident attestation ledger.

  • Drop-in TypeScript SDK — observe mode deploys in minutes
  • Privacy by design: metadata and content hashes only — payload bodies are never stored
  • Detects HTTP 402 payment demands across emerging rails (x402-class), recording raw terms
  • Spend anomaly detection — novel sellers, velocity spikes, payment-terms drift — feeding attested human triage
  • Engineered to never block or slow your agents — a load-test-verified guarantee
  • Batch-attested event ledger with idempotent ingestion

Observe-and-attest available now to design partners; spend mandates, enforcement, and machine-commerce risk intelligence are on the roadmap.

Enterprise Preview

ACA — Autonomous Commerce Assurance

Assurance foundations for agentic commerce across retail, logistics and robotics fleets, healthcare procurement, government, and financial services: a tamper-evident attestation ledger with checkpoint-anchored, cryptographically verified chain integrity as the trust substrate for autonomous transactions.

  • Hash-chained, append-only attestation ledger
  • Checkpoint-anchored chain verification with signed anchors
  • Merkle-rooted, HMAC-signed ledger checkpoints
Enterprise Preview

AIS — Agentic Investigations Suite

Regulated fraud-claim investigations with an append-only attestation ledger and fail-closed attestation coverage. Every investigative action, human or machine, is provably recorded.

  • Reg E unauthorized-ACH intake with AI classification and deterministic fallback
  • Business-day deadline clocks with pre-breach escalation
  • Case briefs with attested evidence checklist tracking
  • Investigator-initiated 45-day extended investigation
  • Entity-clustered alert triage with deterministic priority scoring
  • Evidence briefs with per-item source citations — no uncited assertions
  • Attested human adjudication with enforced override reason codes

Playbook-driven and industry-extensible: banking dispute claims shipped first — healthcare claims, insurance claims, government benefits, and e-commerce chargeback playbooks are on the roadmap.


IONDRA Risk Intelligence

The AI fraud and risk platform IONDRA started with.

Illustrative demo telemetry — not customer traction

Real-time

Fraud detection

6 agents

AI models

AES-256

Encryption at rest

TLS 1.3

In-transit security

your-iondra-host/dashboard

Illustrative demo data

Total Alerts

48

+3 critical

Open Cases

12

Needs review

Flagged TXNs

237

of 18,493

Risk Score Avg

24

Low risk

Transaction Risk Trend (7 Days)

M
T
W
T
F
S
S

Live Alerts

Velocity anomalyhigh
Geo mismatchcritical
Device changemedium
Threat Intelligence

Fraud Is No Longer Just Stolen Passwords — It Is Synthetic Trust.

AI-generated faces, voice cloning, synthetic identities, and manipulated media are changing how fraud happens. IONDRA helps organizations detect anomalies, verify trust signals, and respond before digital deception becomes financial, reputational, or operational damage.

Loading secure video…

Built for banks, healthcare, government, fintech, and enterprises that cannot afford to trust appearances alone.

Deepfake Impersonation

Detect identity manipulation and suspicious behavioral signals before they cause damage.

Synthetic Identity Risk

Expose fabricated or blended identities before onboarding damage occurs.

Transaction & Behavior Anomalies

Correlate fraud signals across users, accounts, devices, and events.

Discover IONDRA Fraud Defense

Schedule a Fraud Defense Discovery Call with our team

Core Capabilities

Every layer of financial risk, covered.

From real-time detection to full investigative case closure — IONDRA handles the entire risk lifecycle.

AI Fraud Detection

Behavioral ML models analyze transactions in real time, flagging anomalies with explainable reasoning.

  • Velocity checks
  • Device fingerprinting
  • Geo-behavioral analysis

Dynamic Risk Scoring

Every entity — user, device, transaction — receives a live risk score built from dozens of weighted signals.

  • Multi-factor scoring
  • Score history timeline
  • Threshold-based alerts

Case Management

A full investigative workflow: triage, assign, document, escalate, and close fraud cases — with complete audit trail.

  • Status workflows
  • Evidence attachments
  • Resolution outcomes (TP / FP / inconclusive)

Alerts & Monitoring

Automated alert pipelines deliver instant notifications via webhooks when thresholds are breached, with severity-based routing.

  • Webhook delivery
  • Severity routing
  • Case auto-creation from alert

Multi-Tenant Intelligence

Strict org-level data isolation with per-tenant fraud rules, watchlists, RBAC, and an independent audit trail.

  • Org-level isolation
  • Per-org fraud rules and watchlists
  • Isolated audit trail per tenant

Compliance & Audit Logs

Every action captured with actor, timestamp, and context. GDPR-aligned, RBAC enforced.

  • Tamper-evident audit trail
  • RBAC access control
  • Export for compliance

What You Get

Four capabilities. One platform.

Every shipped surface, accessible from day one. No add-ons required.

Detect

AI risk scoring on every transaction in real time

Investigate

Full case lifecycle from first alert to resolved outcome

Comply

Tamper-evident audit chain and 5-role RBAC

Integrate

REST API, OpenAPI spec, and onboarding docs

How It Works

From raw data to resolved case.

A closed-loop intelligence pipeline that gets smarter with every transaction.

1

Ingest

Connect any data source — transactions, events, logs — via REST API, webhooks, or streaming.

2

Analyze

AI models score events across configurable risk signals in real time.

3

Detect

Anomalies are identified and classified with explainable AI reasoning.

4

Act

Automated rules trigger blocks, flags, or alerts. Cases are auto-created for human review.

5

Learn

Every human decision feeds back into the model, continuously improving accuracy.

Proof of Intelligence

AI that explains itself.

Every fraud flag comes with a full breakdown — not just a score, but the exact signals that triggered the decision. Give your team the context to act fast and with confidence.

  • Risk score decomposition across all signals
  • Plain-English explanation for each flag
  • Analyst override with feedback loop
  • Decision audit trail for compliance

Transaction #TXN-88421

$4,200 · Merchant: CryptoExchange.io

Risk Score: 82

Why this transaction was flagged

Velocity Anomaly87/100

14 transactions in 3 minutes

Geo Deviation73/100

Login from Nigeria, TX from UK

Behavioral Drift61/100

Unusual merchant category

Device Risk44/100

New device, no prior history

Account Age12/100

Account 3 years old, trusted

AI Recommendation: Block transaction and create investigation case. High confidence fraud indicator based on velocity + geo patterns.

Enterprise Trust Layer

Security you can stake your compliance on.

IONDRA is built for regulated industries where data integrity, access control, and auditability are non-negotiable.

End-to-End Encryption

All data encrypted at rest (AES-256) and in transit (TLS 1.3). Zero plaintext storage.

Role-Based Access Control

Granular permissions from Viewer to Super Admin. SSO and MFA supported.

Multi-Tenant Isolation

Strict org-level data isolation. No cross-tenant data leakage by design.

Tamper-evident Audit Logs

Every action timestamped and recorded. Exportable for compliance and legal review.

Compliance Architecture

Built around SOC 2 (in progress) control families: access control, audit logging, encryption, and incident response. Certification roadmap available on request.

GDPR & CCPA Aligned

Data residency controls, right-to-erasure support, and consent management built in.

Use Cases

Built for every financial context.

Banking

Real-time payment fraud for tier-1 banks

  • Wire fraud prevention
  • Account takeover detection
  • Regulatory reporting

Watch the 90-second overview

Watch: why the agent economy needs a receipt layer — and how IONDRA provides it (90 seconds).

Built on an open standard

IONDRA publishes a public reference-log endpoint for MCA (Machine Commerce Attestation), the open Certificate-Transparency-style evidence standard for the agent economy. Endpoint availability alone is not a log-health signal.

Learn more at iondra.com/mca

API & Integrations

Developer-first by design.

Full REST API, webhook delivery, and SDK support. Integrate IONDRA into any stack in under an hour.

POST /api/transactions/ingest
Bearer auth · Illustrative example
curl -X POST \
https://your-iondra-host/api/transactions/ingest \
-H "Authorization: Bearer iondra_..." \
-H "Content-Type: application/json" \
-d '{
"amount": 4200,
"currency": "USD",
"deviceId": "dev_9f3k2",
"countryCode": "US"
}'
// 201 Created — transaction ingested and scored
{ "id": 1042, "riskScore": "0.8200", "status": "blocked", ... }

Build agents that can prove what they did.

Hash-only receipts in a public log, sealed in signed checkpoints, checked by a verifier anyone can run.

No account required · Open standard · Open-source verifier