VoiceRun
Talk to us
Product

Introducing VoiceRun Edge

Your voice agents. On your infrastructure.

VoiceRun Edge brings the complete platform into your cloud or private data center: code-first agent execution, releases, testing, evaluations, and observability—on infrastructure your team controls. Use local models, external model services, or both.

One-click install. Your infrastructure. Ready to build.
Talk to our teamExplore the platform

Customer environment

Private network · customer-managed access

Edge active

VoiceRun Edge

Control and runtime

Private console
Agent runtime
Releases
Evaluations
Sessions
Telephony

Model services

Local · external · hybrid

Speech-to-textRealtime ASR + turn-taking
Language modelAgent reasoning + tools
Text-to-speechStreaming voice generation
DatabaseLocal or managed PostgreSQL
RecordingsCustomer-controlled storage
AccessPrivate administrative tunnel

Cloud convenience, infrastructure control

The platform comes to the data

Voice is some of the most sensitive data an organization handles. Calls can contain identity, financial details, health information, internal procedures, and the context behind a customer relationship. For many enterprises, sending every turn through a shared SaaS control plane is not an acceptable architecture.

VoiceRun Edge changes the deployment boundary without shrinking the product. Your developers still build agents in Python and work through the VoiceRun CLI. Operators still get sessions, traces, recordings, releases, and evaluations. The difference is where it all runs: inside your project, attached to your network, database, storage, telephony, and security controls—whether that boundary is in AWS, GCP, Azure, or your private data center.

Your infrastructure

Run the platform, models, agent code, recordings, and operational data inside your AWS, GCP, Azure, or private data-center boundary.

Your choice of models

Run speech recognition, language, and text-to-speech models locally, connect external model services, or mix both for each stage of the voice pipeline.

Code-first agents

Build and version Python handlers with the VoiceRun CLI. Agent logic runs as an isolated process beside the platform, not in a shared SaaS runtime.

Production workflow included

Create releases, test live conversations, inspect transcripts and traces, run evaluators, and replay recordings from the private console.

One-click installation

From infrastructure to live agent—in one click

Select your target and launch. VoiceRun's one-click installer brings up the complete, versioned platform in AWS, GCP, Azure, or your private data center—including the runtime, console, and model integrations—without assembling the stack yourself.

  1. 01

    Choose the boundary

    Deploy in AWS, GCP, Azure, or your private data center.

  2. 02

    Install in one click

    Launch the complete approved release inside your network and security controls.

  3. 03

    Connect

    Open the private console and CLI through your secure administrative tunnel.

  4. 04

    Build and test

    Create an agent, run a live session, inspect the trace, and promote a release.

You own the operating boundary

Private by architecture

VoiceRun supplies the versioned runtime and updates. Your team decides where it runs, who can reach it, which providers it can call, and how long its data remains. Normal agent operation does not depend on the VoiceRun cloud.

  • Voice agent source code and release history
  • Call recordings, transcripts, traces, and evaluation results
  • Network policy, encryption, identity, and data retention
  • Model-provider credentials and telephony integrations

VoiceRun Edge

Bring your next voice agent home

Tell us about your cloud, data boundary, telephony, and model requirements. We'll help your team plan the deployment and get the first agent running.

Discuss a private deployment