Changelog
AppService status
Back to All

v253 Genesys webchat integration and 2FA 🛡️

3 days ago by Michał Brzezicki

“You have to go through it all - the good and the bad." - The System, Black Mirror, S4E4 🤖
SentiOne Automate version 253 was released on 19.08.2025. This new release brand new integration of SentiOne Webchat with cloud version of Genesys platform as well as significant improvement to security by adding 2FA.

Genesys

From version 253 onward, SentiOne Automate supports a seamless handover of conversations - from chatbots to live agents in Genesys Cloud. This empowers users to escalate complex or sensitive inquiries to human operators effortlessly, ensuring a smooth transition and improved customer experience.

In a nutshell, the integration allows for:

  • Smooth escalation paths when the bot encounters queries beyond its capability.
  • Improved service quality by enabling human intervention right when it’s needed.
  • Whole conversation in one place - both bot and agent parts of the conversation are visible in webchat widget for the end user with clear information who is the author of each message.
  • No need for additional integrations on website level - webchat snippet code embedded on existing pages are backward compatible. Messages from Genesys agents appear live when they are sent.

Please follow tutorial on how to configure the integration both on Automate and Genesys side

Two Factor Authentication (2FA)

This release introduces significant improvement to security both on SaaS and on-prem environments!

  • 2FA provided by usage of TOTP (Time-based one-time password)
  • Widely available standard supported by Google Authenticator app or other similar applications
  • 2FA can be enabled on environments without access to the Internet
  • Site admins can enforce Organization wide 2FA policy among members
  • Browser and IP is automatically saved for 30 days, so that there is no need provide TOTP at every login
  • Built-in mechanism to prevent brute force attacks

For more details you can on how to use and manage this new feature please read the relevant documentation page

💊 IMPROVEMENTS

  • [LLM] Passing results of tool calling in LLM Say block

🐛 FIXES

  • [NLU] Frontend displayed error when recognized intent no longer exists in training set when analyzing phrase in NLU
  • [Flow] Do not put in project changelog information that NLU was imported if it already existed in project during flow import
  • [LLM Buffering] Bot response buffering split domains which caused incorrect spelling in Google TTS
  • [Dialer] Dialer public API would return HTTP Error code 500 when phone number was invalid
  • [AI Credentials] Fixed validation of new URL formats for OpenAI model deployments in Azure

📚 DOCS


Follow the changelog for more exciting new features! 👋