Contact person: Customizable AI Voice Agent Webapp

Phone:Show

Email:Show

Location: seberang jaya, Malaysia

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I need a small, browser-based webapp that lets a non-technical user spin up and tweak their own AI Voice Agent in minutes. The main use case is appointment setting, but the same interface should leave room for future skills such as upselling and sales follow-ups.

Core workflow
• Front-end: A clean drag-and-drop canvas (think mini flow-builder) where the user can:
 – write or paste the system prompt
 – toggle skills (appointment setter is the default)
 – drop in their Vapi API token and any model parameters
 – type the phone number the agent will call from / to
 – switch on a WhatsApp reminder step

• Back-end: All voice calls and transcriptions run through Vapi. Scenario logic (e.g., sending reminders, branching based on call outcome) should be executed by either Make or n8n—whichever you’re more comfortable with. Please wire the flow so new skills can be added later without re-coding the UI.

• Output for the user: a single “Launch” button that spins up the scenario, tests the line, and reports success or errors in real time.

Tech notes
I’m happy with React, Vue, or Svelte on the front-end; choose what will keep the interface snappy. Node or Python on the server is fine as long as Vapi webhooks are handled securely. Use native Make/n8n modules wherever possible.

Deliverables
1. Source-controlled code for front-end, server, and scenario.
2. A short README showing how to drop in fresh Vapi, Make/n8n, and WhatsApp credentials.
3. Demo video (or loom) proving an appointment is scheduled and the reminder is fired.

Acceptance criteria
• User can sign in, drag blocks, save, and relaunch without page refresh.
• Calls are initiated through Vapi and logged.
• Reminder message lands on WhatsApp within 60 seconds of booking.
• No secrets are hard-coded; all keys loaded from env vars.

If you’ve built anything similar, especially with Vapi, Make, n8n, or drag-and-drop UIs, please link to that past work in your proposal." (client-provided description)