What is the Palmate AI chatbot?
Palmate AI brings customer communication channels into a central AI chatbot platform. It combines LLM-based language understanding, RAG document retrieval and integrated system actions through Function Calling.
The RAG engine finds relevant knowledge before generating a reply. Answers can draw on company documents, product catalogs or approved Q&A, with handoff rules for questions outside scope.
How is an AI chatbot trained?
Training starts with one of three sources: uploaded PDF documents, a website URL or manually entered question-and-answer pairs. The system indexes this content for RAG. When a new question arrives, the chatbot retrieves relevant documents and uses their information to form an answer.
Sample questions help review replies as knowledge sources are updated. Questions without sufficient information can be routed to live support or a ticket workflow.
How does agent handoff work?
When a conversation needs a person, Palmate transfers its full context and message history to an agent. The agent does not need to start the conversation again.
Topic-based routing can also be configured. Some requests are handled by the bot, while actions needing individual assessment go to an agent or ticket system. The same workflow defines when the bot resumes.
Proactive chatbot and channel automations
Palmate can do more than answer incoming messages. Proactive messaging scenarios can be configured for specific pages, user segments or time periods. Product recommendations and cart reminders are planned around data access and customer permissions.
WhatsApp scenarios can include shipping updates, post-order information and satisfaction surveys. Instagram workflows can use post context for comments and messages. Outbound messages follow the channel’s permission and template rules.
Technical setup and integrations
The low-code interface helps configure automation scenarios. Development required for custom workflows and Chat API connections is evaluated against the systems involved.
Enterprise IVR voice bot scenarios using Text-to-Speech and Speech-to-Text can also be reviewed as part of integration. Rollout is planned after channel access, knowledge sources and action permissions have been checked with sample conversations.










































