Tasks That React: Event and Webhook Triggers for AI-Powered Workflow Automation

Scheduled Tasks gave Revel Digital users a way to put the platform on autopilot — run an AI analysis every morning, refresh a data table every hour, audit your fleet every night. Powerful, but it all runs on a clock. The task fires because it's 6:00 AM, not because anything actually happened.

The problem is that most of the work worth automating isn't tied to the clock at all. It's tied to events. A new piece of media gets uploaded. A new device comes online in the field. A ticket gets created in your help desk. A row changes in some upstream system. These moments don't wait for the next scheduled run, and forcing them into a polling loop ("check every five minutes if anything changed") is both wasteful and slow.

So we extended Tasks beyond the schedule. A task can now be triggered three ways:

⏰ On a schedule — the cadence-based behavior you already know.
⚡ By an event — fired automatically when something happens inside your Revel Digital account, like new media being uploaded or a new device registering.
🪝 By a webhook — fired by an HTTP request from anything that can send one, including third-party workflow tools.

Pair any of those triggers with the AI agent task type and you get something genuinely different from a fixed automation: a task that reads the context of what just happened, reasons about it with a prompt you wrote in plain language, and takes action accordingly. Not "if this, then exactly that" — more like "when this happens, here's what I want you to figure out and do."

Here's how it works.

Event Triggers: React to What Happens in Your Account

Event-triggered tasks listen for things happening inside Revel Digital and fire the moment they occur. Two of the most useful event sources to start with are new media uploaded and new device registered — and both pair naturally with an AI agent.

Run agentic workflows on event triggered tasks

Vet every upload automatically

Picture a large account where dozens of people across multiple locations can upload media — franchise managers, regional marketing, agency partners. Every one of those uploads is a small act of trust: you're assuming the content is on-brand, appropriately sized, and safe to put on a public screen. At scale, that assumption breaks.

With a new-media event trigger, every upload kicks off an AI agent task the instant the file lands. The agent receives the details of the uploaded media as its trigger payload and runs whatever review you've described in your prompt. For example:

Review the uploaded image. If it contains nudity, violence, hate symbols, profanity, or anything else unsuitable for a public-facing display, flag it and notify the marketing team. Otherwise, approve it for use.

The same trigger handles a completely different problem — fleet fit. Digital signage fleets are rarely uniform: you've got portrait lobby screens, ultrawide menu boards, and standard landscape displays all in the same account, and a single uploaded asset almost never suits all of them. Instead of asking uploaders to export five versions, let the agent do it:

When new media is uploaded, generate appropriately sized variants for each display resolution in our fleet so the asset is ready to schedule on any screen.

The uploader does one thing — drop in a file. The agent handles the review, the resizing, and the housekeeping, every time, without anyone remembering to.

Onboard new devices the moment they appear

A new device registered event is the start of a small, repetitive onboarding chore: name it sensibly, drop it in the right group, assign a default schedule, maybe tag it by location. An AI agent task on that event can read the device's registration details and take care of the setup — applying your naming conventions, slotting it into the correct group based on its properties, and confirming it's ready to play — so a screen that comes online in the field is provisioned correctly before anyone has to touch it.

Webhook Triggers: Let Anything Start a Task

Event triggers cover what happens inside Revel Digital. Webhook triggers open the door to everything outside it.

Each webhook-enabled task gets its own endpoint URL. Any system that can send an HTTP request can fire that task and hand it a payload — and that includes the entire ecosystem of workflow automation tools your organization already uses: Zapier, Make, n8n, Power Automate, or a custom internal app. If a tool can POST JSON, it can drive a Revel Digital task.

This is where the AI agent really earns its place. A traditional webhook integration is rigid: it expects a specific payload shape and breaks the moment the source changes. An AI agent task is forgiving — it reads whatever payload arrives, interprets it against your prompt, and decides what to do. You're describing intent, not mapping fields.

A few patterns this unlocks:

Incident-driven messaging. Your monitoring or alerting platform posts an incident to the webhook. The agent reads the severity and location and pushes an appropriate notice to the affected screens — an outage banner in one building, business-as-usual everywhere else.
Event-driven content. A registration platform posts the day's schedule each morning. The agent turns it into a session board for the lobby displays, refreshed automatically as sessions change.
Cross-system reactions. A CRM milestone, a completed deployment, a new hire in your HR system — any of these can post to the webhook and let the agent decide what should appear on screen as a result.

Because the payload is just data and the behavior is just a prompt, you can wire up automations we never explicitly built a feature for. The webhook is the connection; the AI agent is the logic.

Why the AI Agent Changes the Equation

It's worth being clear about what the agent adds, because "triggered automation" on its own isn't new — webhooks and event hooks have existed forever.

The difference is the reasoning layer. A conventional trigger executes a predetermined action. An AI agent task receives the trigger context and decides what's warranted, guided by a prompt written in plain language. That means:

One task handles many cases. A single "review uploaded media" task covers offensive content, brand violations, and resizing, because the agent reasons about each upload rather than running a fixed script.
You describe outcomes, not steps. You don't have to anticipate every payload variation or branch. You state what you want, and the agent works out how to get there from whatever it's handed.
Behavior is easy to change. Adjusting the automation means editing a sentence in the prompt, not rebuilding an integration.

Putting It Together

The triggers and the agent compose into a simple model:

A trigger decides when a task runs — on a schedule, on an account event, or on an incoming webhook.
The trigger payload supplies the context — what was uploaded, which device registered, what the third-party system sent.
The AI agent and your prompt supply the judgment and action — reading that context and doing what you asked, in plain language.

Scheduled tasks keep your platform tidy on a cadence. Event triggers make it react to its own activity. Webhook triggers let the rest of your stack reach in and put the platform to work. In every case, the AI agent is the piece that turns a raw trigger into an intelligent action — no rigid integrations, no brittle field mappings, just a description of what you want done.

Get Started

If you're already a Revel Digital customer, event and webhook triggers are available on your AI agent tasks today. Create a task, choose an event source or grab its webhook URL, write a prompt describing what the agent should do, and use Run Now to test it against a sample payload before you turn it loose. Check the knowledge base for setup walkthroughs, or reach out to your account team if you'd like help designing a workflow.

Not a customer yet? This is what it looks like when your signage platform stops waiting for instructions and starts responding to your business as it happens.