Every release, dated and described. Subscribe by booking a demo and we'll keep you posted.
Voice agents now answer your phone line. Inbound calls land in your thread with live transcription and full context. Kantum is also installable as a mobile PWA: approve actions, read alerts, and stay in the thread from your phone, no app store required.
Choose Claude, GPT, or Gemini globally or per message. Switch mid-thread without losing context. No lock-in to one provider.
Build multi-step automations triggered by events across your channels and tools. Each run is logged, and every write inside a flow still passes the approval gate. Full run history per flow.
Agents can now use tools through the Model Context Protocol. Connect MCP-compatible tools and let specialized agents call them inside governed workflows.
Stand up purpose-built agents for specific jobs: reception, support, follow-up, scheduling. Embed one on your website or summon it internally. Each agent runs on the model you choose, coordinated by the main operator.
Email, WhatsApp, Slack, Telegram, and tickets in one stream. Context follows the customer across channels, and Kantum replies in the channel that fits.
Kantum now escalates by severity: an in-app flag, then a text, then a phone call if the urgent item is not acknowledged. You define the thresholds and who is on call.
Persistent memory of your clients, projects, and how you work. Read, edit, export, or delete any entry. Every memory answer cites its source and date.
In the main workflows, write actions are proposed before they land. You set which actions are gated and who can approve, and an audit log records every action your AI agents take.