💭 Knowledge Base (Staging)
- Document uploads - Attach PDFs or docs to improve AI precision and grounding during calls.
 - Scoped access - Knowledge bases are private to each user for security.
 - Fast creation - Create and delete knowledge bases asynchronously in about 3 seconds.
 - Context-aware answers - Agents can reference figures from uploaded files without extra prompting.
 
🔗 Conditional Workflows (Staging)
- Post-call conditions - Send email or SMS only when conditions match, such as status equals interested.
 - JSON field matching - Parse post-call analysis and match fields or search for text.
 - Voicemail detection - Filter voicemail interactions before triggering automations.
 - Fewer external tools - Common MAKE flows are now native for clearer setup and debugging.
 
📞 Web-to-Phone Switching (Staging)
- Seamless handoff - Move a live web chat to a phone call with full transcript and context preserved.
 - Configurable prompts - Define follow-up rules and caller numbers per agent.
 - Multi-channel foundation - Sets up future continuity with SMS follow-ups.
 
📆 Campaign Management (Staging)
- Drafts and scheduling - Prepare, edit, and schedule campaigns before activation.
 - Contacts and CRM - CSV uploads and CRM connectors populate campaign lists.
 - State and filters - Improved state management and activity filtering ahead of beta.
 
💰 $300K+ Annual Savings for One Team
See how healthcare teams are saving $300K a year — and agencies are automating 90K daily calls — all with Fluents. Book a demo and find out exactly what that looks like for your business.

Explore Insights
Read deeper dives into our philosophy, technical breakdowns, and customer success stories.
FAQs about this week’s update
Quick answers to the most common questions from this release.
You can scope uploaded documents per user and attach a knowledge base to any agent that needs it for targeted answering.
Post-call analysis can be parsed as JSON and matched on fields such as status equals interested. You can also add voicemail and pickup conditions.
Yes. The agent preserves the full transcript and recent context when switching the conversation to a phone call.




