Hot platform updates of the week

Weekly News Digest

We continue to actively develop the platform's beta version, enhancing stability, expanding functionality, and improving performance. In today’s digest, we cover the latest updates from the past week and share our plans for future development.

New booking tools

We are expanding the platform's functionality and introducing booking and appointment scheduling tools for agents:

You can now configure the available schedule by specifying responsible staff members and time slots, view and confirm or reject appointments, and manage your list of services and their prices. Service names and prices can be localized into Russian and English.

To enable these tools, don't forget to add the new functionality in your agent settings:

New experimental connectors

We have introduced new experimental connectors for your AI agents. You can now create bots for Discord and several other social networks. This is particularly useful for community management, including within the gaming industry. The bot can handle direct commands and also analyze message history:

New functionality for dialogue analysis

The analytics section now displays problematic messages that your AI agent was unable to answer. There, you can work with the agent to update the knowledge base based on the full conversation history, helping to prevent similar issues in the future:

With this functionality, we have laid the groundwork for the intelligent augmentation of knowledge bases, transforming our platform from a mere tool for managing agents into a comprehensive corporate knowledge hub.

Here, you can also view security events and track conversations involving the transmission of personal data or instances of attempted malicious use.

Archiving leads / requests

You can now archive old requests and view the archive. This improves usability when dealing with a large number of requests:

Platform optimization

We are continuing to optimize the core, focusing on performance speed and resource consumption. Token usage is now even more efficient thanks to caching, and security models are running faster than ever.

The next step involves stabilizing MCP servers and improving tool usability, enhancing documentation and examples, introducing even more new tools (including tools for game developers), as well as refining the interface and adding more AI agent templates.