Voice AI that thinks and talks like your best rep
Build AI phone agents that handle real conversations. Sub-300ms latency, natural turn-taking, and full observability. Connect your backend logic via webhook or MCP. We handle the rest.
200 free minutes/month. No credit card required.
// 1. Create an assistant
curl https://api.voicerail.ai/v1/assistants \
-H "Authorization: Bearer $APLOYEE_KEY" \
-d '{
"name": "SupportBot",
"system_prompt": "You are a calm, competent support rep.",
"personality": {
"formality": "neutral",
"assertiveness": "medium",
"empathy": "medium"
},
"reasoning_mode": "webhook",
"webhook_url": "https://your-domain.com/aployee"
}'
// 2. Handle reasoning in your backend
app.post("/aployee", async (req, res) => {
const { utterance, state } = req.body;
if (utterance.includes("balance")) {
const balance = await lookupBalance(state.userId);
return res.json({
reply: `Your balance is ${balance}. Anything else?`,
state: { ...state, lastBalance: balance }
});
}
return res.json({ reply: "Can you tell me more?", state });
});
// 3. Make a call
curl https://api.voicerail.ai/v1/calls \
-H "Authorization: Bearer $APLOYEE_KEY" \
-d '{ "assistant_id": "asst_xxxx", "to": "+15551234567" }'Everything you need for production voice AI
Purpose-built infrastructure for real-time AI conversations. No telephony expertise required.
Sub-300ms Latency
Real-time responses that feel natural. Our dual-agent architecture keeps dialogue flowing while your backend reasons.
AI-Native Architecture
Purpose-built for LLMs. Conversation agent + reasoning agent work in parallel for human-like interactions.
Your Logic, Our Voice
Connect your backend via webhook or MCP. We handle speech, you handle business decisions.
Zero Telephony Complexity
Inbound numbers, outbound dialing, transfers, recordings. No carriers, no SIP, no configuration.
Natural Turn-Taking
Barge-in detection, intelligent interruptions, dynamic pacing. Conversations that feel human.
Personality Controls
Fine-tune formality, assertiveness, empathy, and tone. Predictable behavior without black-box psych models.
Full Observability
Real-time timeline, waveform visualization, latency breakdown across ASR, TTS, and reasoning.
Adaptive Reasoning
Built-in reasoning modes from fast to deep thinking. Let the AI adapt to conversation complexity.
How Aployee works
A dual-agent architecture that keeps conversations flowing naturally
Caller connects
Inbound or outbound, we handle the telephony. Your number, our infrastructure.
AI converses
Our conversation agent maintains natural dialogue while your backend reasons asynchronously.
You decide
Your webhook receives context and returns business decisions. Results blend naturally.
Built for your use case
From sales to support to custom voice products
Outbound Sales
AI reps that qualify leads, handle objections, and book meetings. Your CRM provides the intelligence, we handle the call.
Inbound Support
Agents that authenticate, lookup accounts, and resolve issues. Complex problems route to your backend for resolution.
Voice-First Products
Building a voice product? Skip telephony complexity. Focus on your logic via webhooks while we handle everything else.
Ready to build voice AI?
Get started in minutes. Try the live demo or dive into the docs. 200 free minutes every month.
Simple, transparent pricing
Pay for what you use. No hidden fees. No surprises.
Free
- 200 minutes/month
- 2 assistants
- 7-day retention
- Webhook reasoning
- Community support
Pro
+ $29/project/month
- Unlimited assistants
- 30-day retention
- MCP integration
- Priority scaling
- Priority support
Enterprise
- Volume pricing
- Custom SLAs
- 90-day+ retention
- Dedicated support
- Compliance add-ons