VoiceLLM is a full-stack AI communication platform developed over a 4-week build cycle. The system delivers three integrated AI platforms- an AI Inbound Caller, AI Outbound Caller, and an AI Chatbot. Each platform is designed to automate customer interaction using advanced voice and language models. Built with Next.js + TypeScript, the project includes a custom admin dashboard, public landing page, and a scalable backend powered by Supabase. The platform integrates Twilio for telephony, ElevenLabs for natural-sounding speech synthesis, and GPT models for conversational intelligence, alongside custom internal tools provided by the client. A full custom Figma design system was created for the UI, and Resend powers automated email communication. The result is a polished, production-ready ecosystem for managing and deploying AI-driven voice and chat interactions.