Quickstart

This quickstart follows the dashboard flow for getting one agent ready for real Users. It assumes you already have access to a Hyponema workspace.

1. Open the workspace

Open the dashboard and select the correct workspace from the sidebar. Home shows the workspace overview and is the fastest place to notice missing setup, no workspace access, or billing interruptions.

2. Add provider credentials

Go to Settings and open Providers. Add the model, speech, telephony, or messaging provider credentials your agent needs. Keep provider credentials in the dashboard or your backend systems; never place them in browser code.

Use API keys only for backend-to-Hyponema calls. Treat workspace API keys as secrets and rotate them when access changes.

3. Create an agent

Go to Agents and create an agent. The agent detail page uses these visible tabs:

• Basics for identity, system prompt, opening message, language, and runtime basics.
• Knowledge for attaching workspace knowledge sources.
• Actions for integrations and post-session work.
• Controls for availability, listening behavior, and personalization fields.
• Test for validating the agent before launch.
• Distribute for Web widget, API, Phone, WhatsApp, and Webhooks setup.
• Privacy for agent-level privacy controls.

Older route names such as models, voice, variables, listening, publish, deploy, and advanced are not visible agent tabs in the current dashboard.

4. Add knowledge and actions

Use Knowledge to browse workspace documents and attach the relevant sources to the agent. Use Actions to connect provider-backed integrations, create custom actions, and choose which actions the agent can use.

5. Test before launch

Use the agent Test tab for a focused preview, and use the top-level Tests page for reusable regression coverage. The top-level Tests page has Tests, Example sets, Judges, and Runs.

6. Distribute the agent

Open the agent Distribute tab and choose the channel:

• Web widget for website embeds.
• API for backend-owned text or voice integrations.
• Phone for PSTN calls.
• WhatsApp for messaging deployments when enabled for the workspace.
• Webhooks for outbound event notifications to your backend.

Copy setup values from the dashboard instead of hard-coding undocumented URLs or payloads.

7. Operate after launch

Use Conversations for conversations, Users for people and their records, and Calls for PSTN sessions. Use Memory and Observability from Monitor for workspace-level memory visibility, traces, live monitoring, and logs.

Use Billing to review wallet credits, rate-card meters, credit ledger, payment method, top-ups, and auto-recharge.