Activepieces Integration

Use your Asyntai assistant inside Activepieces flows

Paid Plan Required: The Activepieces integration works on the Starter, Standard, and Pro plans. View pricing

Overview

Activepieces is an open source automation tool. The Asyntai piece sends a customer message to your assistant and returns the answer. You then route that answer anywhere in the flow: an email reply, a Slack message, a helpdesk ticket, or a database row.

Available Actions

The Asyntai piece provides two actions:

Ask Assistant - Sends a customer message to one assistant and returns the answer
Custom API Call - Calls any Asyntai API endpoint with the same connection

Note: The piece installs from npm. You need a self-hosted Activepieces instance, or platform administrator rights on Activepieces Cloud.

How to Install

Follow these steps to connect Asyntai to Activepieces:

Get Your API Key

Open the API Settings page and generate a key:

https://asyntai.com/settings/api/

Copy the key and keep it safe. Anyone with the key can reach your Asyntai account through the API.

Install the Asyntai Piece

In Activepieces, go to Settings, then My Pieces, then Install Piece. Type this package name:

@asyntai/piece-asyntai

Connect Your Account

Add an Asyntai step to a flow. Create a new connection and paste your API key. Activepieces checks the key at once and shows an error if it is wrong.

Build the Flow

A flow needs two parts:

Trigger: any Activepieces trigger, such as a new email, a form entry, or a webhook
Action: Asyntai - Ask Assistant

Map the customer message and a stable conversation ID from the trigger into the Asyntai action.

Send the Answer Back

Add one more step to deliver the reply: Gmail, Slack, a helpdesk, or an SMS service. Use the answer field from the Asyntai step as the reply text.

Test and Publish

Run a test with a real customer question. Check the answer, then publish the flow.

Ask Assistant Inputs

Assistant Website (required) - the assistant that answers. The list loads from your account.
Customer Message (required) - the text the assistant must answer.
Conversation ID (required) - a stable value for this customer or thread, such as an email address or a ticket number.
Source Channel (required) - where the message arrived. Email channels add email instructions to the assistant.

Ask Assistant Outputs

answer - the reply from the assistant.
conversation_id - the thread the reply belongs to.
ai_disclosure - the AI disclosure text. Show it to the customer.
success - whether the call worked.

Tip: Improve the answers by adding instructions, FAQs, and knowledge base content in your Asyntai Dashboard.

Conversation History

Asyntai keeps the history under the Conversation ID you send. Use the same value for the same customer, and the assistant remembers earlier messages. Use a new value to start a fresh thread.

Each call adds a message to the history, so a repeated run creates a duplicate entry. Keep this in mind when you set up retries.

Example Use Cases

Email support: a new email arrives, the assistant drafts the answer, and Gmail sends it
Slack helpdesk: a question in a Slack channel gets an answer posted back in the thread
Form reply: a contact form entry triggers an answer by email within seconds
Ticket triage: the assistant answers the simple tickets and leaves the rest to a person
SMS answers: connect a text message service and answer customers on the phone

Source Code

The piece is open source under the MIT licence. The package name on npm is: