Prevent losses earlier
Catch risky refunds, vendor changes, payments, and system actions before they turn into write-offs or remediation work.
Loading…
Straight answers
Your answers for reducing execution risk
See what Decionis gives your team before high-stakes work moves forward: fewer preventable losses, faster approvals people can trust, and signed proof for every decision that needs to stand up to finance, risk, and audit.
Approve with confidence
Give every approver the policy, evidence, owner, and verdict they need to allow, hold, or escalate.
Keep proof ready
Hand finance, risk, compliance, and auditors one signed record instead of asking teams to rebuild the story later.
Direct answers
The questions that come up first
Written to be forwarded — so finance, risk, operations and compliance can all get the same answer about what Decionis does and where it sits.
What is Decionis?
Decionis is the independent Execution Authority for high-stakes actions: deterministic infrastructure that decides whether an exact action has authority under your organisation's own policy, binds an ALLOW or named-human approval to execution, and records the observed outcome in an Ed25519-signed Decision Dossier. Access is not authority. Intent detection is not authority. Approval is not authority until it is bound to execution. The commit-bound policy check answers allow, escalate, or block in under 120ms (p95 decision compute). Your team gets one verdict before the action happens, with signed evidence that can travel to finance, risk, audit, and the workflow owner.
Read the quickstartWho is Decionis for?
Decionis is for high-volume commerce and Shopify Plus retail merchants, plus finance, risk, compliance, operations, and product teams that need execution authority for checkout margin protection, payments, refunds, vendor changes, credit decisions, banking operations, procurement, hospitality workflows, trading controls, HR approvals, or AI-driven actions.
See the team and companyWhy doesn't approving every AI agent action work?
Because at volume, approving stops being a decision. Anthropic measured that Claude Code users approve 93% of permission prompts, and shipped classifiers specifically to reduce permission fatigue (Anthropic Engineering, 25 March 2026). A separate 40,000-run study by Scale X found people missed 1 in 3 genuine threats while clicking through, and 7% approved every prompt they saw. That leaves two bad escapes: approve everything by hand and get nothing done, or skip permissions entirely and hope. Decionis takes the prompt off the common path — your policy answers deterministically in under 120ms, and a named person is pulled in only when your rules require one, so the approvals someone does see are rare enough to deserve attention.
How a gate replaces the promptHow do I stop babysitting my AI agent without giving it free rein?
The two defaults are approve-everything and skip-permissions, and neither is a control. Policy-based gating is the third option: your policy is evaluated on every consequential action before it commits, so routine work proceeds without asking and only what your rules flag reaches a person. The policy is authored and versioned by your team and evaluated at runtime rather than compiled into one application, and every answer leaves a signed record anyone can verify. You stop reviewing each command and start reviewing the policy.
Put a gate in the agent loopWho gave Decionis the authority to stop an action?
You did. Decionis holds no policy of its own. It evaluates your organisation's policy — authored, approved, and versioned by your team — against each intended action at runtime, and when your rules require a person, a named human in your organisation approves. Decionis carries no opinion about what should be allowed; it applies the decision you already made and proves afterwards which version of your policy was in force.
How execution authority is delegatedIsn't this just logic my own engineers could write?
Any team can write a check inside one service. What is expensive is everything around it: policy authored and versioned outside the application, evaluated identically across every surface an action can leave from, maker-checker approval so the author cannot be the approver, a named human in the loop when the rules demand one, and a signed record anyone can verify afterwards. Decionis is the permission step as infrastructure, not business logic living inside a single app.
See what execution authority is notHow do I keep a record of what my team decided without anyone maintaining it?
Records that depend on someone writing them down decay — teams abandon architecture decision records within weeks, and the products built on that habit have repeatedly failed. Decionis records a decision as a byproduct of authorizing an action: every evaluation writes a row and a signed Decision Dossier, and every policy change is a dated, attributed revision showing what it replaced. A month of running the gate leaves you a decision record whether or not anyone maintained anything.
Why the record writes itselfHow does Decionis protect margin at checkout?
Decionis runs an inline, cost-basis-aware margin floor at checkout. In under 120ms it evaluates each order against your verified unit cost and blocks coupon-stacking and price drift that would ship below your net-margin floor, then seals a signed Decision Dossier for every held order so finance and dispute defense have non-repudiable proof of what was held and why.
See the margin-governance comparisonWhy choose Decionis over ServiceNow IRM or IBM OpenPages?
Choose Decionis when you need confidence before a high-stakes action happens, not only documentation after the fact. ServiceNow IRM and IBM OpenPages are strong for broad GRC programs, issue tracking, and control management. Decionis is built for the execution moment: it evaluates the action, returns one verdict, and leaves signed proof before money moves or systems change.
Compare with ServiceNow IRMWhat tradeoffs should you weigh when choosing Decionis versus ServiceNow IRM or IBM OpenPages?
Decionis is narrower than a full GRC suite by design. ServiceNow IRM and IBM OpenPages can manage broad programs, policies, issues, and risk workflows. Decionis is the execution authority your team uses when the priority is preventing the wrong transaction, workflow step, or agent action from moving forward and proving why the decision was made.
Review the OpenPages tradeoffHow is Decionis different from Guardrails AI, Lakera, or LangChain Guardrails?
Guardrails AI, Lakera, and LangChain Guardrails help shape model behavior. Decionis is the execution authority around the business action that follows. It checks policy and evidence before a real workflow, payment, customer action, or system change proceeds, then records the decision in a signed Decision Dossier.
Review guardrail comparisonsHow does setup work?
Most teams start with one workflow in shadow mode. Decionis connects to a representative signal source, maps the decision context, and shows what would proceed, pause, or escalate before you enforce anything. From there you can add routes, promote policy versions, and generate validation packs as confidence grows.
See setup stepsWhich systems does Decionis integrate with?
Decionis works where your teams already approve, route, and execute work: Shopify, BigCommerce, Slack, Microsoft Teams, Power Automate, Salesforce, ServiceNow, Google Cloud, Vertex AI, Jira, ERP systems, and API or webhook-driven workflows.
Review the evidence indexCan Decionis convert existing policy documents into executable policy packs?
Yes. The policy draft endpoints extract rules from unstructured sources — PDFs, JSON, XML, transaction and audit logs, uploads — and from live policy-source connectors for Jira, Confluence, Notion, Google Drive, Git repositories, and SAP. Every synthesized rule is validated through the rules compiler before it is returned, and clauses that cannot become executable predicates come back as review items for a human, never as broken rules.
Read the policy encoding docsWhat is a Decionis validation pack?
A validation pack gives reviewers the broader evidence bundle around your Decionis workspace: audit metadata, integrity hashes, and regulator-facing proof that complements each per-decision dossier.
Open validation pack docsWhat is a Decision Dossier?
A Decision Dossier is the signed proof your team gets for a high-stakes decision. It records what was attempted, what evidence was checked, which policy applied, who owned the decision, and why Decionis recommended proceed, hold, or escalate.
See Decision ObservabilityWhat is the Decionis authority lifecycle?
The Decionis authority lifecycle turns one decision into memory your organization can reuse. A Decision Dossier captures the moment, a Decision Chain links related decisions, and an Execution Ledger preserves the record alongside overrides, policy drift, and policy evolution.
See the execution mapHow does Decionis handle overrides and policy drift?
Decionis keeps overrides visible instead of letting them become hidden workarounds. The original Decision Dossier remains intact, the exception is attached as an Override Dossier, and repeated patterns show where policy should evolve.
Open override docsHow does Decionis work with Microsoft Teams and Power Automate?
Decionis keeps Teams as the collaboration surface and Power Automate as the workflow rail while policy stays in Decionis. Microsoft flows call the shipped policy-check and trigger path, Decionis resolves the Decision Graph, and the resulting Decision Dossier routes back into the Teams surface for review, escalation, and verification.
Review the Microsoft surfaceHow does Decionis fit into a Google Cloud or Vertex AI stack without losing governance or human approval?
Decionis fits as a thin Vertex AI Extension or ADK tool surface that forwards agent intent into Decionis Protocol for the actual certified decision. The adapter can verify Google identity at the edge, return one certified verdict plus Decision Dossier proof, emit verified_decision usage metadata, and preserve deterministic policy checks, human approval paths, org-wide and per-decision kill switches, and audit-ready evidence while Google Cloud and Vertex AI stay in the intelligence and distribution layer.
Open the Vertex AI quickstartHow does Decionis stay cloud-agnostic while Microsoft Teams and Power Automate remain workflow surfaces today?
Teams and Power Automate are workflow surfaces, not the permanent home for policy. Decionis stays cloud-agnostic because the execution-time decision boundary remains in Protocol, while Google Cloud, Vertex AI, Microsoft, ERP, and other systems can act as distribution or execution surfaces around the same Decision Dossier, approval, and validation-pack proof model.
See the cloud-agnostic operating modelCan Decionis work with Google Cloud or Vertex AI while preserving the current Microsoft Teams, Power Automate, and Decision Dossier proof story?
Yes. Decionis can keep Microsoft Teams and Power Automate as current workflow surfaces while a Vertex AI Extension or ADK tool path sends Google-side intents into Decionis Protocol. The Decision Dossier, verification envelope, validation pack, human-approval posture, and enterprise hardening path for PSC, regional posture, KMS-backed deployment controls, and Google-shaped usage export stay aligned across both surfaces.
Review the fitCan Decionis support regulated or compliance-sensitive workflows?
Yes. Decionis is built for regulated workflows where auditability and policy enforcement matter before execution. Public materials reference EU AI Act Article 12, SEC/FINRA-style audit trails, Central Bank execution provenance, signed Decision Dossiers, and downloadable validation packs.
Review governance postureDoes Decionis allow autonomous execution?
Yes. Decionis supports autonomous execution with policy gates, while still pausing, blocking, escalating, or deliberately restraining gray-zone decisions when evidence, confidence, or policy is insufficient.
Review operating modesIs Decionis legitimate and trusted for high-stakes execution?
Yes. Public trust signals include Microsoft ISV Partner status, NVIDIA Inception participation, public docs and quickstart flows, security and governance materials, Decision Dossier verification, validation packs with integrity hashes, and org-wide and per-decision kill switches.
Review public trust postureWhat are the limitations and tradeoffs?
Decionis is not a full GRC suite, generic workflow builder, or developer-only LLM guardrail layer. It is built for one high-value job: helping your team know which high-stakes actions should proceed, pause, or escalate before the action becomes expensive.
Review governance postureWhy would a CFO choose Decionis?
A CFO chooses Decionis to reduce preventable loss, shorten approval cycles, and keep audit-ready proof around spend, approvals, and risk-bearing operations before money moves.
See CFO fitWhat is Decision Observability?
Decision Observability means your team can explain why a high-stakes action happened, who owned it, and what proof supported it. Decionis starts with a verifiable Decision Dossier, links related outcomes into a Decision Chain, and preserves the institutional record in an Execution Ledger.
Open the category pageStart where the business case is easiest to see.
Pick a workflow where one wrong action is expensive: refunds, payments, vendor changes, ERP actions, or regulated customer updates. Decionis shows what would proceed, what would be held, and the proof your team gets before you enforce anything.
See Shadow Mode