<aside> 💡 This release introduces Google Gemini support and improves omnichannel delivery logic.
</aside>
Version 3.3.0 introduces support for Google Gemini via Vertex AI, providing a robust alternative for AI chat interactions alongside OpenAI. This enables us to use completely EU-based data centers. We have also refined our omnichannel capabilities: the system now prioritizes a user's specific channel preference (such as WhatsApp) before falling back to the default configuration hierarchy, ensuring messages reach users where they are most responsive.
Support for Google Gemini LLM AD-977
Added integration for Gemini via Vertex AI, including support for RAG (Retrieval-Augmented Generation) and function calling.
Enhanced channel prioritization logic AD-995
Campaigns now attempt delivery via the user's preferred channel first, falling back to the system hierarchy only if delivery fails.
Localized Excel templates AD-982
Admin users now receive Excel templates corresponding to their selected language setting.
Clearer warnings for outdated survey versions AD-983
The editor now clearly prompts users to update to the current version and prevents accidental saving of old drafts.
Automatic phone number formatting during upload AD-996
Dots and other non-standard characters are now automatically stripped from phone numbers during contact file uploads.
WhatsApp button responses as text AD-987
When a user clicks a WhatsApp button, the selection is now recorded as a text response.
Increased character limits for statistics AD-990
Expanded storage capacity for campaign names and contact degree fields in survey statistics.
Campaign emails now include WhatsApp variables AD-984
Fixed an issue where notification emails regarding started campaigns were missing variable data.
Fixed publishing error for support bots AD-985
Resolved a validation error preventing the publishing of support bot surveys when modules were bypassed.
Cleaned up test survey popups AD-988
Test survey confirmations no longer display empty WhatsApp variable fields.
Configuring Gemini (Vertex AI)
To utilize Gemini, the 'Vertex AI Service Agent' IAM role must be assigned to the service account. Additionally, update the survey configuration JSON to include the 'aiagent' key referencing the desired agent ID.