Platform · Agents
Conversations that do work
An agent on Fluents is voice, prompt, and skills together. Configure once, then run the same agent on inbound lines or inside outbound campaigns, with a clean path to your team when automation should stop.
Agents that finish the job
Look up orders, book calendars, update CRM, and hand off without losing the thread. More than a voice demo.
Operators and builders together
Operators configure day-to-day in the dashboard. Builders extend with REST actions and webhooks. Quality comes from turn-taking and brand voice. Outcomes come from skills wired to your stack.
Agent capabilities
Concrete tools under the conversation, organized for operators who need to ship.
- Configurable agents: Prompt, voice, model, and skills in one dashboard. Start from a library or write custom instructions.
- Skills that act: End call, transfer, DTMF, external REST actions, appointment scheduling, knowledge search, and SMS or email, mid-call or as follow-ups.
- Natural conversation: Smart turn-taking, barge-in, tunable endpointing and speed, noise cancellation, multilingual and custom voices.
- Warm transfer with context: Warm-transfer to humans with the caller on the line and full context when policy or complexity says stop.
Intelligent routing
Understand intent, send callers to the right queue, or resolve inline when the job is contained.
Systems in the loop
Look up orders, book calendars, push CRM updates, and call your APIs so containment means completion.
Always-on coverage
Answer every inbound line without staffing for the worst peak. Elastic capacity underneath.
How teams stand up an agent
Define the job
Order status, booking, qualification. Write the outcome the agent owns, and when it should escalate.
Attach skills & systems
Connect calendars, CRM, ecommerce, or your APIs so the agent can look up, write back, and complete work mid-call.
Test, then put on a number
Run test calls from the dashboard, tune prompts and dials, then link inbound lines or campaign steps.
Improve agents from real call data
Insights and transcripts feed the next prompt and skill change. The same loop runs for inbound lines and outbound campaigns.
- Review transcripts and insights: See how the agent handled the job, where callers dropped, and when handoffs fired.
- Tune prompts and skills: Update instructions, knowledge, and mid-call actions from what the reports show.
- Redeploy and compare: Ship the change to inbound lines or campaign steps, then measure the next batch of calls.
For builders
Test calls, prompt libraries, mid-call actions, and event subscriptions when the UI isn't enough.