Patchstore24Shopify
Scenarios
One intent. One Scenario.
Full control.
A Scenario is one intent's resolution path: conditions your team sets, actions your team allows, all scoped to a single customer intent. Not one monolithic agent you have to trust.What a Scenario can check and do
Scoped to one intentChecked against live dataApproved actions onlyHolds when unmetTested per intentDeployed one at a timeScoped to one intentChecked against live dataApproved actions onlyHolds when unmetTested per intentDeployed one at a timeScoped to one intentChecked against live dataApproved actions onlyHolds when unmetTested per intentDeployed one at a time
Check order stateVerify fulfillment windowsExecute approved actionsEscalate with contextLog every decisionExpand intent by intentCheck order stateVerify fulfillment windowsExecute approved actionsEscalate with contextLog every decisionExpand intent by intentCheck order stateVerify fulfillment windowsExecute approved actionsEscalate with contextLog every decisionExpand intent by intent
76%+
AI resolution rate
End-to-end resolution on paths pre-approved by your team
100%
Accurate on what it acts on
A Scenario fires only when it is certain, and escalates the rest
24h
To first value
Start with your highest-volume intent and expand from there
Trusted by customers on
Wicked ClothesShopify
Zendesk Marketplace reviewZendesk
Anatomy of a Scenario
Conditions your team sets. Actions your team allows.
Every Scenario has the same shape: an intent it belongs to, conditions checked against live data, and an explicit list of allowed actions. Here is each part of that shape, shown in a real conversation.
Before any Scenario runs, the Customer Intent Engine classifies the conversation. Only the Scenario for that intent is eligible to fire. Every other automation stays out of the conversation entirely, so behavior never bleeds across intents.
Customer
Hi, when will order #5127 arrive? I need it before Friday.
Aide
Order #5127 shipped yesterday and is on track to arrive Thursday. Here is the live tracking link so you can follow it to your door.
✓Intent detected: order status · Order status Scenario fired · No other Scenario eligible
A Scenario's conditions are verified in your systems at the moment of the conversation. Was the order placed recently? Is the fulfillment window still open? Does the customer record match? Every condition is checked before Aide acts, not assumed.
Customer
I just placed order #8843 and realized it is going to my old address. Can you change it to 41 Cedar Lane?
Aide
Done. Order #8843 has not entered fulfillment yet, so I updated the shipping address to 41 Cedar Lane. You will get a confirmation email with the new address shortly.
✓Conditions met: placed 2 hours ago · fulfillment window open · address verified
Each Scenario carries an explicit list of allowed actions: reply to the customer, execute a change in your systems, escalate to a person. When it acts, it acts inside that list. There is nothing it can reach for that your team did not put there.
Customer
Please cancel order #2461, I ordered the wrong size and want to reorder.
Aide
Cancelled. Order #2461 had not shipped, so I cancelled it and refunded $46.80 to your original payment method. It should appear within 3 to 5 business days, and you can reorder any time.
✓Action executed: order cancelled in Shopify · Allowed by this Scenario
One unmet condition is enough to stop the action. Aide holds, tells the customer what happens next, and escalates to your team with its findings attached. The person picking it up starts with the full picture, not a blank ticket.
Customer
Order #7305 arrived but the jacket was missing from the box. Send it or refund me please.
Aide
I checked order #7305: the packing record shows all three items scanned into a single box, which does not match a missing-item claim on its own. I have escalated this to our team with the packing and weight records attached, and someone will follow up within one business day.
✓Condition unmet: packing record conflicts · Held and escalated with findings
Before a Scenario ships, the Agent Simulator replays it against your real past conversations for just that intent. Your team watches exactly how it would have handled each one, adjusts the conditions or actions, and approves it. One Scenario, tested on its own, with nothing else in the mix.
Customer
I want to return order #1198. The fit is not right.
Aide
No problem. Order #1198 is within the 30-day return window, so I have emailed you a prepaid return label. Your refund of $62.40 will process as soon as the carrier scans the package.
✓Simulator replay · Returns Scenario · 180 past conversations · Awaiting team approval
Rollout is per Scenario, along the Gradual Automation Pathway. Start with your highest-volume intent, watch it perform in production, then approve the next one on evidence. You are never asked to turn on an agent. You turn on a Scenario.
Customer
Any update on order #6674? Tracking has not moved in two days.
Aide
Order #6674 left the regional facility this morning after a carrier delay, and delivery is now expected Wednesday. The tracking link will update within the hour.
✓Order status Scenario · Live on its own · Returns Scenario in review
Aide is the agentic AI platform for customer experience. Scenarios are its unit of governance: control lives at the intent, so you decide what runs per Scenario, not wholesale. See how each one is enforced in agent governance or what agents do with them in AI agents.
Customer
Hi, when will order #5127 arrive? I need it before Friday.
Aide
Order #5127 shipped yesterday and is on track to arrive Thursday. Here is the live tracking link so you can follow it to your door.
✓Intent detected: order status · Order status Scenario fired · No other Scenario eligible
The Agent Governance Engine
Governance built into every Scenario
Because every automation is a Scenario, control is per intent. The Agent Governance Engine enforces each one exactly as your team approved it.Tested per intentThe Agent Simulator replays each Scenario against your real past conversations for its intent. Nothing ships until your team has watched it work and approved it.
Fires only when certainA Scenario runs when every condition is met and confidence clears your threshold. Anything short of that goes to a person with context already attached.
Every run in the Action TraceThe Action Trace records each run: the intent detected, the conditions checked, the action taken, and why. Auditable per Scenario, and across your operation.
The next Scenario, suggestedThe Continuous Learning Engine watches what your team still resolves by hand and suggests the next Scenario worth building, with the evidence behind it.
Getting started
Works with your existing stack
Conditions read live data where it lives, and actions execute in the systems you already run. One click into your helpdesk, direct actions in your systems of record. No migration, no replatforming, no engineering.
Resolves and updates tickets inside your existing workflows.
Closes conversations end to end, right in the shared inbox.
Automates the macros and rules your team already relies on.
Reads live orders and executes changes, holds, and refunds.
Order data and actions, handled natively on your store.
Grounds every answer in the customer record you trust.
Frequently asked questions
A Scenario (formally, Agentic SOPs) is one intent's resolution path: the conditions your team sets and the actions your team allows for a single customer intent. When the Customer Intent Engine detects that intent, the Scenario checks its conditions against live data and, only if every one is met, executes an approved action. It is the unit of governance in Aide: testable in isolation, auditable per intent, and deployed only when your team approves it.
Anything the decision depends on: the intent detected, live order data, tracking and fulfillment status, whether a time window is still open, and customer fields from your connected systems. Conditions are verified at the moment of the conversation, against your systems of record, not against cached assumptions.
Only the actions your team allowed for it. That can be replying to the customer, executing a change in your systems such as a cancellation, an address update, or a refund, or escalating to a person with context attached. A Scenario cannot take an action that is not on its list.
The Scenario does not fire. Aide holds the action, tells the customer what happens next, and escalates to your team with its findings already attached. A held Scenario is the system working as designed, not failing.
A monolithic agent is one behavior you have to trust wholesale: change anything and you have changed everything, and when it misbehaves you cannot isolate why. Scenarios cut the problem per intent. You can test one Scenario in isolation, audit one intent's runs in the Action Trace, and change one resolution path without touching the rest.
In the Agent Simulator. It replays the Scenario against your real past conversations for that intent, so your team sees exactly how it would have handled each one: the conditions checked, the action it would have taken, and the reply it would have sent. You adjust, approve, and only then does it run in production.
One Scenario at a time, along the Gradual Automation Pathway. Most teams start with their highest-volume intent and see first value within 24 hours. From there you expand on evidence: watch a Scenario perform, approve the next, and change what runs on its own at any time.
Get a demo for pricing.
Procedures and deployment, defined
The vocabulary of deploying automations safely, defined.
- Agent operating procedure (AOP)
- Agentic SOP (ASOP)
- AI agent action
- AI agent workflow
- Business logic for AI agents
- CX workflow automation
- Decision tree vs agentic procedure
- Deterministic vs probabilistic agent execution
- Intent-scoped deployment
- Natural-language workflow authoring
- Resolution playbook
- SOP for AI agents
Automate one intent at a time, on your terms
Conditions your team sets, actions your team allows, and evidence before every expansion. That is what a Scenario gives you.