Gemini AI integration
Replora uses Google Gemini, accessed through Vertex AI, as its language intelligence layer: review replies, call conversations, complaint analysis, and business insights. Only the content needed for each task is sent to the model. This page explains where Gemini is used and what data it receives.
Where Gemini is used
Four places: generating review replies (AI Responder), holding phone conversations and producing transcripts summaries (AI Receptionist), writing apology messages and classifying feedback (Review Booster), and analyzing patterns to produce action cards and coach answers (AI Business Insights).
What data is sent
Only task content and configuration: review text and tone settings for replies; conversation audio-derived text and your knowledge base for calls; feedback text for classification; aggregated review/call/complaint text for insights. Replora does not send account credentials, and does not use your data to train models for other businesses.
Quality controls
Star-based routing and approval workflows act as human checkpoints on model output. You can edit any AI draft before it posts, and tone settings constrain style per location.