# Webhook Inbound

Send a POST request to a single endpoint whenever something happens in one of your other systems, and your AI employee picks it up and acts on it in the background. You get a 202 response immediately; there is no waiting for a reply. The same employee that answers you in chat can react to a failed deploy, a new order, or a support ticket the moment it fires.

Webhooks push events from your AI workforce to your systems in real time. When an employee completes a task, your project management tool gets notified. When an approval is needed, your Slack channel gets a message. When a scheduled execution finishes, your data pipeline picks up the output. Events flow out automatically, no polling required.

Configure which events trigger webhooks and where they go. Route task completions to one endpoint, approval requests to another, and error notifications to a third. Each webhook includes a full payload with context: what happened, which employee did it, what the output was, and relevant metadata.

Webhooks are especially valuable for building custom automation pipelines. Chain an AI employee's output into a downstream system: employee researches leads, webhook notifies your CRM to update records, CRM triggers an email sequence, email sequence results webhook back to the employee for follow-up analysis. Closed-loop automation with your AI employee at the center.

## Push Agent Events to Your Systems in Real Time

Webhooks let Sistava push events to any HTTP endpoint you control as they happen. Task completed, approval requested, employee hired, credit threshold crossed, error detected. No polling, no scheduled fetches, no stale data. Your external systems know about events the moment they occur.

This is the standard pattern for production integrations between SaaS platforms and your internal infrastructure. When an AI employee finishes a research task, your project management tool gets updated. When an approval is needed, your ticketing system creates a ticket. When credits run low, your finance system gets an alert. All automatically, in real time.

## Configure Which Events Trigger Webhooks

Not every system needs every event. Webhook subscriptions are configurable per event type and per delivery endpoint. You can send task completion events to one URL and approval requests to a different URL, or fan out all events to a central event bus for routing downstream.

Event payloads are structured JSON with consistent schema per event type. Each payload includes event type, timestamp, relevant entity IDs, and enough context to act on the event without a follow-up API call in most cases. Full payload schemas are documented in the API reference.

Webhook deliveries include retry logic for failed deliveries (non-2xx responses or timeouts) with exponential backoff. Delivery history is visible in the developer dashboard so you can inspect successful and failed deliveries, view the exact payload sent, and manually replay a delivery for debugging.

## Webhooks Enable Event-Driven AI Workforce Automation

Combining webhooks with the REST API creates a fully event-driven automation loop. An AI employee completes a task, the webhook fires, your system processes the result and uses the API to start the next phase, which may involve spinning up another AI employee or sending results to a human reviewer. This pattern is how sophisticated organizations build multi-step automated workflows on top of Sistava.

For developers building products on top of the platform, webhooks are the mechanism for giving end users real-time updates without building polling infrastructure. Your product subscribes to relevant events, routes them to your database and notification systems, and delivers a responsive user experience backed by Sistava's agent workforce.

## How It Works

**External systems send events directly to your AI agent via a dedicated inbound webhook URL, and the agent acts on them immediately.**

Every AI employee gets a personal webhook endpoint. When an external system, a CRM, an e-commerce platform, a monitoring tool, or a custom app, sends a POST request to that URL, the agent receives the payload and begins processing it as a new task. No polling, no manual triggers, no intermediary queue. The event lands and the agent acts.

Configure the webhook in your external system once and it works indefinitely. Your AI agent reads the incoming payload, extracts the relevant fields, and decides what to do based on its skills and duties. A support agent can receive a new Zendesk ticket and draft a reply. A data agent can receive a raw export and process it into a report. Inbound webhooks turn your AI employee into a live endpoint your systems can push work to.

## Use Cases

### E-commerce team triggers AI on new orders

Every new order fires a webhook to the AI employee, which confirms inventory, sends the customer a personalized update, and flags exceptions.

### Support team activates AI on ticket creation

A new support ticket fires a webhook. The AI agent classifies it, drafts a response, and routes it to the right team in seconds.

### Dev team notifies AI on CI/CD events

A failed build fires a webhook to the AI employee, which investigates the error, checks recent commits, and posts a summary to Slack.

### Marketing team responds to form submissions instantly

A lead form fires a webhook. The AI agent enriches the lead, scores it, and queues a personalized follow-up, all before the rep refreshes their screen.

## Comparison

| Before | After |
|---|---|
| External events require a human to notice and respond. | Webhooks trigger the AI agent instantly, with no human in the loop. |
| Connecting external tools to the AI workforce requires custom code. | Any tool that sends webhooks can activate an AI employee immediately. |
| Event-driven workflows require a middleware layer to build. | The AI agent listens for webhooks natively, no glue code needed. |
| Response time to critical events depends on who is online. | The AI employee reacts to every event the moment it fires. |

## FAQ

### How do I verify that a webhook came from Sistava?

Every webhook delivery includes an HMAC signature in the X-Sista-Signature header, computed using a secret you set when creating the endpoint. Verify this signature in your handler before processing the payload to confirm authenticity.

### What happens if my endpoint is down when an event fires?

Sistava retries failed deliveries with exponential backoff over 24 hours. If delivery ultimately fails, the event is marked as failed in the delivery log and you can manually replay it once your endpoint is restored.

### How many webhook endpoints can I configure?

You can configure multiple endpoints per workspace, each with independent event subscriptions. There is no fixed limit on the number of endpoints, though very high fan-out configurations may be subject to rate limits on delivery volume.

### Can I filter webhook events by employee or team?

Yes. Webhook subscriptions can be scoped to specific employees or teams in addition to event types. This lets you set up targeted integrations, like routing events from your sales AI team to your CRM and events from your research team to your knowledge base.

### Can external systems send events directly to my AI agent?

Yes, each AI employee can accept inbound webhooks, so any external service can push data or trigger work in real time. You configure the endpoint once and the agent acts on incoming payloads automatically.

> We hooked our e-commerce platform directly to our AI agent. Every order above a threshold triggers the agent automatically and it handles the review without us touching it.
>
> Amara D., Founder · early-stage startup

## Where Webhook Inbound fits

Webhook Inbound is part of How you plug them into your systems.

Build on your AI workforce with a full REST API, client SDKs, and webhooks in both directions. Expose any employee as an MCP server for Claude, Cursor, and Windsurf, or as an A2A agent for cross-platform agent collaboration. Everything the UI can do, the API can do.

- [How you plug them into your systems](/en/features/developer): Programmatic access to your workforce.

## Read the guide

- [Guide: Webhook Inbound](/en/guide/channels/webhooks)

## More in Developer API

- [REST API](/en/features/developer/channel_api): Every AI employee is reachable over a REST API: send a prompt to POST /api/v1/chat with the employee's ID and get their response back as JSON, running through the same execution pipeline as web chat with full access to that employee's skills, tools, and memory. Requests are synchronous, blocking for up to five minutes, and authenticate with a Bearer API key shared across every programmatic channel (REST API, MCP Server, A2A, Webhooks, Email). The channel is off by default and must be enabled from Settings > Technical > Channels before it accepts requests.
- [Let Other AI Clients Use Your Employee](/en/features/developer/channel_mcp): Connect Claude Desktop, Claude Code, Cursor, Windsurf, ChatGPT, or any Model Context Protocol client directly to your AI employees. Generate an API key, add one config block, and your coding tools can send prompts to any employee by ID and get a real response back. No custom integration code required.
- [Let Other AI Agents Work with Your Employee](/en/features/developer/channel_a2a): Turn your workforce into an agent that other AI agents can call directly. Any client that speaks Google's A2A protocol can discover your employees, hand them a task, and get the result back, all without touching your web UI or writing REST API glue code.
- [Build with the Sistava SDK](/en/features/developer/channel_sdk): A planned client library for JavaScript, Python, and other runtimes that wraps the existing employee REST API. It is on the roadmap, not published: today you call employees from code through the REST API guide directly, and the SDK would only make that same integration point faster to reach.
- [Outbound Webhooks](/en/features/developer/outbound_webhooks): Connect a webhook URL, your own backend, a Zapier or Make scenario, a Slack incoming webhook, anything that accepts an HTTP POST, and your employee gets a tool it can call on its own during a task. It sends a JSON payload to that URL, so the rest of your stack finds out the moment the work is done instead of you checking manually. Because the call sends data outside Sistava, each send waits for your approval by default, the same as an outbound email.

## Explore

- [Every feature](/en/features)
- [Hire an AI employee](/en/market)
- [Pricing](/en/pricing)