Voice Atlas Overview
Voice Atlas centralizes all the conversational knowledge you want to manage through voice, chat, and web experiences. These docs explain the core concepts so you can design, build, and maintain Atlases with confidence.
Rather than simply deploying static information, Voice Atlas lets people interact with your knowledge using AI that understands natural language, follows policy, and routes answers across every channel. Built-in responsible AI guardrails mean there is no prompt engineering to tinker with—the assistant stays polite, appropriate, and refuses to return hallucinated responses even when users probe for edge cases.
What Is an Atlas?
An Atlas is a curated knowledge base that powers the answers your assistants deliver. Each Atlas combines:
- Content items: Questions, answers, and optional API responses that provide dynamic data.
- Channels: web chat, Slack®, Microsoft Teams™, and other endpoints where the Atlas can respond.
- Insights: Analytics and feedback that reveal how people are interacting with the content.
Who Should Use These Docs?
- Administrators rolling out Voice Atlas across an organization.
- Content editors responsible for building Atlases and keeping information fresh.
- Developers integrating Voice Atlas responses into custom applications.
Next Steps
- New to the platform? Follow the Onboarding Guide.
- Ready to build? Jump to the Atlas Workspace for hands-on instructions.
- Need a term definition? Browse the Voice Atlas Glossary.
- Need personal assistance? Contact the Voice Atlas support team.
Chatlas™ Web and Mobile Bots
Chatlas™ is the embeddable assistant for deploying text- and voice-enabled bots backed by your Atlas on any website or mobile web experience. Drop the Chatlas™ widget into your pages to deliver conversational answers in a familiar chat UI that mirrors the Voice Atlas guardrails without additional infrastructure.
Pro tip: We require you to whitelist the domains where Chatlas™ is allowed to load. This prevents other sites from embedding your assistant, impersonating your brand, or siphoning proprietary answers.
Responsible AI Platform
Voice Atlas is built for responsible AI deployments, combining advanced conversational tooling with safeguards that enterprises and research organizations require. Every Atlas inherits the following guardrails:
- Ethical AI practices: Content is moderated, tuned to reject harmful instructions, and guided by policy-aware prompts so assistants stay polite, accurate, and free from hallucinations.
- Data privacy and security: Customer data flows through encrypted channels, is isolated per tenant, and is stored under least-privilege access controls that satisfy enterprise and science-grade security reviews.
- Regulatory alignment: Governance workflows, audit trails, and configurable retention policies help teams meet frameworks such as SOC 2, HIPAA, and GDPR while keeping model behavior transparent.
- User safety and transparency: Clear provenance indicators, configurable disclaimers, and conversation logs make it easy to explain why an answer was returned and to trace any AI-driven interaction end to end.
These measures ensure Voice Atlas delivers compliant, trustworthy answers out of the box—no prompt engineering or risky experimentation required.