SignalWire Voice AI Developer Toolkit
APIs, SDKs, and infrastructure for building powerful voice and AI applications with SignalWire.
All fields are required
APIs, SDKs, and infrastructure for building powerful voice and AI applications with SignalWire.
Build conversational agents that can speak naturally, trigger backend actions, and adapt mid-call. With modular prompts, skills, and integrated voice control, you define exactly how your agents behave. Use it end-to-end or integrate just the pieces you need into your existing workflow.
Build intelligent agents that act, adapt, and remember.
Use it for:
Prompt orchestration, tool calls (skills), barge-in handling, and agent memory
How it works:
Use lifecycle hooks, modular skills, and integrated voice stack to design live AI behavior
Use it for:
Ensuring tone control, fallback handling, and reusable logic
How it works:
Define reusable named prompts inside the Agent SDK or via builder tools
Use it for:
Splitting functionality (e.g., routing, billing, support) between roles
How it works:
Use the Agent SDK runtime to register and switch between agents on the fly
Use it for:
Agent memory, personalization, and retrieval-augmented grounding
How it works:
Use Datasphere via Agent SDK to store context and query it in tools or prompts
Core capabilities for provisioning, connecting, and scaling real-time voice apps.
Use it for:
Purchasing and assigning numbers to projects
How it works:
Uses the provision method via the REST API to search, buy, and configure numbers
Use it for:
Adding voice to dashboards, CRMs, or web portals
How it works:
The Call Widget uses call flow routing and the connect method to initiate live calls
Use it for:
Branded in-app calling or embedded voice clients
How it works:
Use the Browser SDK to register users (via the register method), manage sessions, and handle events
Use it for:
Creating internal users or agents with SIP/WebRTC
How it works:
Subscribers are addressable resources that authenticate clients and define call permissions
Use it for:
Auditing, analytics, or BI dashboards
How it works:
Uses the list calls method to retrieve structured records for analysis
Use it for:
Quota tracking, account reviews, internal reporting
How it works:
Use the list calls method with filters to group usage by number, user, or service
Use it for:
Budgeting, chargebacks, or customer insights
How it works:
Retrieve CDRs via the list calls method, then apply custom pricing logic
Use it for:
Sharing usage insights with stakeholders
How it works:
Use the REST API to gather data, then pass results to a PDF generator (e.g., Python, JS)
Use it for:
Managing parity across dev/staging/prod
How it works:
Script provisioning of numbers, subscribers, and call flows using the provision method
Use it for:
Connecting on-prem systems or SBCs to SignalWire
How it works:
Configure SIP trunks, then use the connect method to route calls
Use it for:
Sending and receiving calls over the public telephone network
How it works:
Use the connect method to route between PSTN numbers and SignalWire logic
Use it for:
Messaging over WhatsApp with business-verified numbers
How it works:
Connect approved WhatsApp accounts and manage sessions via the Messaging API
Route, manage, and enhance live calls dynamically.
Use it for:
Dashboards, summaries, or compliance monitoring
How it works:
Add the transcribe method to the call leg before routing to enable live transcription
Use it for:
Dashboards, summaries, compliance.
How:
Add to call leg before routing.
Use it for:
Multilingual support or tone shifting in live conversations
How it works:
Replace or augment live audio using the translate method during the call
Use it for:
Managing long hold times with scheduled callbacks
How it works:
Use logic in your call flow to offer and trigger callbacks during queueing
Use it for:
Real-time analysis, virtual agent participation, or AI-generated audio
How it works:
Attach an AI Agent using the stream method to listen, respond, and act in the conversation
Use it for:
QA, compliance, redaction, or playback
How it works:
Start and stop recordings mid-call using the record method and conditional logic
Use it for:
Smarter routing based on human vs. machine detection
How it works:
Use the amd method to classify pickup type before continuing the call flow
Use it for:
HIPAA, PCI, and custom data privacy enforcement
How it works:
Combine region-aware routing with selective recording and redaction logic
Use it for:
Voice-driven navigation or hands-free experiences
How it works:
Use the gather method with speech recognition to collect input and trigger actions
Use it for:
Real-time ML pipelines, custom analytics, or raw audio capture
How it works:
Use the tap method to stream audio to a WebSocket endpoint
Use it for:
Mid-call updates like starting/stopping recordings or dynamic transfers
How it works:
Use the rpc method via Agent SDK or the REST API to push updates to a live session
Use it for:
Real-time agent routing, escalation triggers.
How:
Analyze speech via sentiment model and adapt call logic.
Let us know how we can help you bring your ideas to life.