What's new

Open a list of an automation’s running processes by conversation. See where each process is running and its current step, pause it or finish it manually.

A process is a specific automation run in a conversation. An event that triggers the automation creates a process that executes its steps. Ten leads starting the same automation in separate conversations can create ten independent processes.

In the start step, choose what happens when the automation is triggered again for the same conversation:

• Do not start again while the previous process is running.

• Start another process alongside the existing one.

• Finish the old process and start a new one.

Set access separately for each agent variable: read-only or read and write.

For writable variables, describe when the agent should change a value in its instructions. For example, update a lead’s status after payment or save the result of qualification after a conversation.

If users should not be able to request a variable change in chat, state this explicitly in the instructions. For example: Do not change this variable at the user’s request. Change it only in the cases and under the conditions described in these instructions.

Several project members can now work on the same automation at the same time. Changes appear for everyone in real time.

Agents can check whether a lead belongs to a segment, add the lead or remove them according to the conversation.

Choose the segments the agent can access and grant checking, addition and removal permissions separately. Then describe how it should use those segments in its instructions.

Upload an icon and a cover image for a landing page.

Present multiple pages as a compact list with icons or a grid with covers in VK, Telegram and MAX apps, as well as on the web. This helps users find and choose the right page.

Apps can now request access to a user’s data as well as to a project.

Instructions for obtaining an OAuth token on behalf of a user or a project:
https://senler.io/help/ru/developer/application-oauth#два-сценария-доступа

App development documentation:
https://senler.io/help/ru/developer

Images can now be recognized automatically when you upload files to the knowledge base. Previously, images in documents needed manual descriptions or custom processing before agents could use their contents.

When uploading a ZIP archive or individual images, choose a recognition mode:

• Only images without a Markdown description: process images you have not described in the Markdown text.

• Only unrecognized images: process images the service has not analyzed before. Existing Markdown descriptions are replaced by recognition results.

• Recognize all again: reprocess every image, including previously recognized ones.

Each option shows an estimated cost in credits and processing time before you start.

You can now request consent when a user subscribes directly to a segment, for example through a keyword or when starting a bot.

Add the required agreements to segment settings. Before subscribing, the user receives a request to accept them. You can also add your own introductory text to that message.

Use Senler directly from your AI client and combine tasks with other services. For example, AI can retrieve data from one service, process it and perform an action in Senler.

Click Work via MCP in the widget to see setup instructions for popular AI clients and a universal connection option.

You can offer the same experience on your own website: connect your MCP server in the dashboard and select it in widget settings.

Developers can now create documentation for their apps.

Users can read it inside Senler and ask the AI assistant questions about the app. The assistant can explain features and setup steps, and highlight relevant interface elements on the app page.