# Keep Your AI. Replace What’s Underneath.

Your AI agent works. The telephony layer underneath it adds latency, cost, and failure modes that do not scale.

2,000+

companies in production

< 1.2s

typical AI response latency

2.7B

minutes processed annually

$0.16

per minute, AI processing

# The Vendor Chain Problem

## Your AI Is Only as Fast as the Slowest Vendor

### Every vendor adds latency

Your stack routes audio through telephony, STT, LLM, and TTS providers. Each hop adds 100–300ms. Four hops turn a fast LLM into a slow conversation.

### Every vendor adds a failure mode

When the call drops at 2am, who gets the ticket? Debugging across four dashboards is the real cost.

### State splits across systems

Conversation context lives in webhooks, Redis, and vendor callbacks. Transfers lose context because no single system owns the call state.

### Costs compound through the chain

Each vendor marks up the layer below it. A four-vendor stack means four margins stacked on top of carrier costs.

## Build a Voice AI Agent

Python

TypeScript

Go

Java

Ruby

PHP

Perl

C++

C#

```python
from signalwire import AgentBase
from signalwire.core.function_result import FunctionResult

class SupportAgent(AgentBase):
    def __init__(self):
        super().__init__(name="Support Agent", route="/support")
        self.prompt_add_section("Instructions",
            body="You are a customer support agent. "
            "Greet the caller and resolve their issue.")
        self.add_language("English", "en-US", "rime.spore:mistv2")

@AgentBase.tool(name="check_order")
    def check_order(self, order_id: str):
        """Check the status of a customer order.
        
        Args:
            order_id: The order ID to look up
        """
        return FunctionResult(f"Order {order_id}: shipped, ETA April 2nd")

agent = SupportAgent()
agent.run()
```

## Multi-Vendor Stack vs. One Platform

### Current: Multi-Vendor Stack

- Telephony provider for numbers and PSTN
- Separate STT and TTS providers
- LLM provider with token billing
- Your glue code connects all four systems
- Every vendor adds latency and cost

### SignalWire: One Platform

- Phone numbers, SIP, PSTN, media streaming included
- STT and TTS run inside the AI processing engine
- One invoice, sub-second latency
- Your AI logic stays the same
- Replace telephony first, consolidate AI when ready
- AI executes inside the media stack

# Migration Path

## Move at Your Own Pace

### Phase 1: Replace telephony

Point your existing system at SignalWire for phone numbers and PSTN. Your AI orchestration stays unchanged. Immediate cost savings and lower latency.

### Phase 2: Consolidate AI processing

Move STT and TTS onto the platform. Eliminate separate provider SDKs and invoices. Reduce network hops and latency.

### Phase 3: Full platform

Move your agent logic to the Python SDK. One stack, one vendor, sub-second latency. Scoped prompts and tools per conversation step.

## Sigmond Runs on SignalWire

A voice and video AI agent built on the SignalWire SDK, wired to a live knowledge base, able to see what you show him. Build the same thing on voice, phone, WhatsApp, or SIP, or an entire white-label platform with SignalWire as the network underneath.

## Keep your AI. Replace what holds it back.

Better latency, lower cost, and a path to the full platform when you are ready.