Six stages from the first description to a reviewed call: create, build, add knowledge, test, launch and review. Every screen below is from hiify, not a stand-in.
Describe the agent in plain language or start from a template. hiify.ai drafts an editable flow for you to review. The published canvas — not Business notes — defines what callers experience.
The draft remains editable. Change the language, steps, voice, and connected tools before it reaches a caller.
Arrange greetings, questions, branches, knowledge answers, booking actions, and CRM steps on one canvas. The route through the conversation is inspectable rather than buried in a prompt.
Warnings stay visible in the builder so missing content or connections can be resolved before publishing.
Add files, web pages, or pasted text for opening hours, services, policies, and frequently asked questions. Knowledge steps draw from those workspace sources.
Storage is metered by plan, and the workspace shows what is ready and how much of the allowance is in use.
Run the actual voice and chat widget inside the builder. The test view lists the flow, tools, and unresolved setup beside the conversation.
Use those checks to connect missing apps, add knowledge, and republish after edits. Testing is part of the build loop, not a final reveal.
Launch on the web widget, or connect your own Twilio account and assign a number for phone calls. Home tracks live agents, conversations, booked appointments, and usage.
The populated workspace keeps operational results beside the setup state, so the team can see what is happening without opening a separate analytics tool.
Open a completed conversation for the caller’s intent, answered questions, actions taken, booked appointment, and the full transcript.
That context makes the next edit obvious: improve a knowledge source, change a branch, or tighten the handoff, then publish the revised flow.
Start with a template or describe the job in plain language. Review every step before you publish.
