Skip to main content

Decoda RWA GuardAvailable for pilot evaluation

Protect the Infrastructure Behind Tokenized Finance.

Decoda RWA Guard brings security monitoring, evidence-grounded investigation, and policy-controlled response workflows together for tokenized financial infrastructure.

Detect threats.Investigate with evidence.Respond under policy.

Illustration with sample data: a Decoda RWA Guard incident moving from detection through investigation, policy review and human-authorized response to a sealed evidence package.
  • Read-only monitoringGuard reads chain activity and never holds customer keys.
  • People authorizeAI and policy recommend. Your team decides.
  • Verifiable evidenceTamper-evident packages an auditor can check offline.
  • One Decoda identityInvitation-only access with shared sign-in and MFA.

The problem

A Valid Blockchain Transaction Can Still Be Unauthorized.

A transaction can be correctly signed, executed and final — and still fall outside the approvals, operational policies and risk limits your institution runs under. The chain verifies the transaction. It cannot verify your authority.

Tokenized-asset operations need more than transaction visibility. Security teams have to see what changed, judge it against policy, act under control — and show afterwards what was decided.

Sample admin transactionIllustration

Valid on-chain

  • Signature verifiedYes
  • Executed without revertYes
  • Finalized on-chainYes
But was it authorized?

Authorized by your institution

  • Approved by the required people?Unknown
  • Within operational policy and limits?Unknown
  • Expected by treasury operations?Unknown

The ledger records what happened — not whether it should have.

  • Privileged access abuse

    Admin keys and privileged roles can be used in perfectly valid transactions while acting outside the authority they were granted.

  • Unauthorized administrative changes

    Role grants, upgrades, pauses and parameter changes can take effect without the approvals your governance requires.

  • Suspicious asset transfers

    Mints, burns and transfers can settle exactly as written and still fall outside expected operations.

  • Operational policy violations

    Limits, allowlists and change windows live in policy documents — not in the contract that executes the transaction.

  • Investigation and evidence gaps

    After an incident, teams must reconstruct what happened and who decided what — and then show it to auditors and oversight.

Introducing Decoda RWA Guard

One Security Operations Platform for Tokenized Assets.

Guard brings monitoring, detection, investigation, policy-controlled response and verifiable evidence into one workspace — so every decision stays connected to the evidence behind it.

Monitor

Continuous monitoring

Watches the contracts, wallets and privileged roles you register on supported EVM networks. Missing or stale data is reported as missing — never as safe.

Detect

Threat detection

Typed detections for privileged activity, suspicious transfers and policy violations, routed by severity into alerts.

Investigate

Incident investigation

Related alerts are correlated into incidents, each with a forensic timeline and a case file your team works from.

Assist

Evidence-grounded AI assistance

Optional AI triage summarizes an incident from its evidence record and suggests predefined runbooks. It recommends; it never acts.

Decide

Deterministic security policies

A rule-based policy engine evaluates every recommended response. The same input always gets the same decision.

Authorize

Approval-controlled responses

Approval quorums and step-up MFA before a response proceeds. Production execution is off by default on every plan.

Prove

Verifiable evidence

Evidence packages carry a SHA-256 manifest and a Merkle integrity root, so any change is detectable offline.

One connected record

Every step links to the one before it.

  1. Asset
  2. Monitoring target
  3. Telemetry
  4. Detection
  5. Alert
  6. Incident
  7. Response
  8. Evidence
See how Guard works

How Guard works

From First Signal to Verifiable Evidence.

AI recommends, a deterministic policy engine decides what needs approval, and people authorize. Every step leaves a record.

  1. 01Detect
  2. 02Investigate
  3. 03Review
  4. 04Respond
  5. 05Preserve evidence
  1. 01Detect

    Continuous monitoring of what you register.

    Guard watches the contracts, wallets and privileged roles in your monitoring scope and raises typed detections for privileged activity, suspicious transfers and policy violations.

    • Privileged-role and admin-change monitoring
    • Severity-based alert routing and escalation
    • Missing data is reported as missing — never as safe
  2. 02Investigate

    Context assembled, not hunted for.

    Related alerts are correlated into one incident with a forensic timeline. Where AI triage is enabled, it summarizes the evidence and suggests predefined runbooks — grounded in the incident record and unable to act on its own.

    • Incident case file and forensic timeline
    • AI-assisted summaries, recommend-only
    • Incident playbooks
  3. 03Review

    Deterministic policy. Human approval.

    A deterministic policy engine evaluates each recommended response and routes it to the people with authority to approve it. Each decision is recorded against who made it and when.

    • Rule-based, repeatable policy decisions
    • Approval quorums and step-up MFA
    • Role-based workspace access
  4. 04Respond

    Response stays under your control.

    Your team carries out approved actions through its own processes and signers. Production execution is off by default on every plan, pilots are recommend-only, and Decoda never holds customer private keys.

    • Slack, webhook and email notifications
    • Recommend-only during pilots, enforced server-side
    • No custody of customer keys
  5. 05Preserve evidence

    Evidence an auditor can verify.

    Evidence packages bundle the original artifacts with a SHA-256 hash per file, a manifest hash and a Merkle integrity root, so any change is detectable offline. Where a deployment has provisioned Ed25519 signing, packages also prove their origin.

    • Tamper-evident evidence packages
    • Offline verification instructions in every package
    • Append-only, hash-chained audit log

Product experience

The Workspace Your Security Team Operates In.

Nine connected screens take an analyst from posture to proof. The previews are illustrations with sample data; in the product, every screen shows only what your workspace actually reports — and says so when data is missing.

Dashboard

Security posture at a glance.

Monitoring coverage, open alerts and incidents, and what needs attention next — derived from canonical runtime facts, not from a frontend guess.

  • Posture strip built from runtime status
  • Open alerts and incidents by severity
  • No data is shown as no data — never as healthy

Illustrative interface · sample data

Evidence-grounded AI

AI That Explains Its Findings.

Guard’s AI-assisted triage works from the incident record — telemetry, detections, transactions and timeline — and points to the evidence behind every statement. It drafts; it recommends. Deterministic policy and your people decide.

  • Evidence-grounded investigationsSummaries are built from the incident record and point back to the evidence they used.
  • Contextual security analysisTelemetry, detections, transactions and the incident timeline are read together, not in isolation.
  • Incident summariesA draft an analyst can review, edit and keep in the case file.
  • Recommended actionsSuggestions are limited to predefined runbooks your team already knows.
  • Deterministic policy evaluationWhether a response needs approval is decided by rules, never by a model.
  • Human approval requirementsEvery recommendation needs a person to approve it. Approving records a decision; it executes nothing on its own.
Illustration with sample data: an AI-assisted incident summary that cites numbered evidence, recommends a predefined runbook, and waits for a deterministic policy decision and human approval.

AI-assisted triage is optional and enabled per deployment. It has no tools that act on-chain, and approving one of its recommendations records a decision — it executes nothing on its own.

Institutional controls

Built for Security Teams That Need Control.

Guard is designed to observe, investigate and record — not to hold your keys or act on its own. Every control here is implemented today, and the limits of a pilot are enforced server-side, not left to a setting.

ImplementedIn the product today — not a roadmap item.

  • Read-only monitoring

    Guard’s JSON-RPC access is read-only: it never calls a signing or transaction-sending method, and Decoda never holds customer private keys.

    Implemented
  • Human-controlled actions

    AI and policy recommend; people authorize. Production execution is off by default, and pilots are recommend-only — enforced server-side.

    Implemented
  • Role-based access

    Workspace roles — owner, admin, analyst and viewer — scope what each person can see and change.

    Implemented
  • Multi-factor authentication

    Shared Decoda sign-in with MFA, mandatory for every member of a pilot workspace, and step-up verification before approvals.

    Implemented
  • Policy enforcement

    A deterministic engine decides which responses need approval; every refused execution is written to the audit log.

    Implemented
  • Evidence integrity

    A SHA-256 hash per file, a manifest hash and a Merkle integrity root make any change detectable offline.

    Implemented
  • Audit history

    An append-only, hash-chained audit log per workspace records governance actions, incidents, responses and exports.

    Implemented

The pilot boundary

What a pilot can — and cannot — do.

During a pilot, your team can

  • Observe blockchain activity and monitor contracts
  • Detect threats; create alerts and incidents
  • Run investigations, including AI-assisted triage
  • Simulate policies and playbooks
  • Approve or reject recommendations
  • Export evidence and review audit logs

Guard will refuse to

  • Sign or broadcast a transaction
  • Execute a state-changing contract call
  • Pause a contract or freeze an asset
  • Move funds, mint or burn
  • Change ownership or admin roles
  • Use a customer wallet key or signer

Every refusal is written to the hash-chained audit log. Decoda never holds customer private keys.

The Decoda ecosystem

The Flagship of the Decoda Ecosystem.

Decoda builds security and operational infrastructure for tokenized finance. Guard is available for pilot evaluation today; Vault and Assets are earlier-stage products that show where the platform is heading.

Flagship security platformAvailable for pilot evaluation

Decoda RWA Guard

Security monitoring, investigation and controlled response for tokenized financial infrastructure.

  • Continuous monitoring and alerts
  • Investigation and AI-assisted triage
  • Policy-gated, human-authorized response
  • Tamper-evident, verifiable evidence
Digital asset operationsIn development · Testnet

Decoda Vault

Prepare, simulate and approve digital asset operations under deterministic policy before anyone signs.

Runs on public testnets only. Not a custodian; never holds private keys.

Learn about Decoda Vault
Tokenization sandboxSandbox prototype · Testnet

Decoda Assets

Model the lifecycle of a tokenized real-world asset — from origination to servicing — in a testnet sandbox.

Synthetic assets and investors on public testnets. No securities are offered or issued.

Learn about Decoda Assets

Pricing

Start with a Scoped Pilot.

Evaluate Decoda RWA Guard against your own infrastructure, then move to recurring security operations when you are ready. Every plan is arranged with the Decoda team — there is no automatic checkout.

Recommended starting point

Scoped evaluation

Pilot

Tailored

Tailored scope and commercial terms

Validate Decoda RWA Guard against your own infrastructure with a defined scope and a guided evaluation.

  • Defined monitoring scope — 1 workspace, up to 5 monitored contracts
  • Guided evaluation with the Decoda team
  • Alerts, incident investigation and evidence workflows
  • Recommend-only response workflows — no production execution
  • Feedback sessions and technical validation
Request a pilot

Recurring security operations

Scale

From $999/ month

Final pricing depends on scope

For teams moving beyond pilot evaluation to recurring security operations.

  • Defined limits — 3 workspaces, up to 25 monitored contracts
  • Continuous monitoring, alerting and incident workflows
  • Investigation workflows and incident playbooks
  • Unlimited evidence packages and audit-ready exports
  • Support scoped to your agreement
Contact sales

Institutional programs

Enterprise

Custom

Custom pricing

For institutions with tailored requirements, a broader monitoring scope and their own security and governance reviews.

  • Tailored institutional requirements and monitoring scope
  • Custom monitoring and integrations, where supported
  • Security and governance requirements reviewed with your team
  • Custom onboarding and support arrangements
Contact sales

On every plan, Decoda recommends and your people authorize. Production execution is off by default and is enabled only for an Enterprise tenant through an explicit, audited agreement. Decoda never holds customer private keys.

Compare plans in detail

Secure Your Tokenized Infrastructure with Decoda.

Explore a scoped evaluation of Decoda RWA Guard with your institution’s security and operations team.