About the service
A platform for assembling support agents: the visual Botpress Studio, a natural-language builder, knowledge bases and tables for structured data, escalation to a human, and the Botpress Desk helpdesk inside the same account. An agent ships to webchat, WhatsApp, Slack and other channels, and integrations pull data from Jira, HubSpot or Salesforce. Metering is not per seat and not per message but per conversation — any exchange with at least two end-user messages — and when the quota runs out a conversation pack is added automatically, with auto-recharge that cannot be switched off. There is one important nuance about model spend: since 14 May 2026 LLM inference, embeddings and web search are bundled into the conversation allotment as a proportional quota, yet workspaces created before that date keep the earlier scheme where AI spend was invoiced separately at provider cost. So for any ready account you should establish when the workspace was created — that decides whether you get one bill or two.