PERSONAL PRODUCT / AI

AI Receptionist & Voice Assistant Platform

AI · Voice · Automation

Personal In Development

Offline-oriented AI voice automation platform exploring automated reception, speech processing and business communication workflows.

DjangoDRFChannelsCeleryRedisOllamaWhisper.cppPiper

The AI Receptionist & Voice Assistant Platform is a personal R&D project exploring how AI voice systems can support business reception and communication workflows.

The project investigates speech recognition, language-model processing, voice responses and backend automation, with particular interest in locally deployable or offline-capable components.

CHALLENGE

Reception workflows often involve repetitive conversations: answering common questions, capturing information and directing enquiries to the appropriate next action.

Building an AI alternative requires coordinating speech recognition, AI processing, voice generation, application state and asynchronous workloads while maintaining acceptable response behaviour.

SOLUTION

The platform architecture combines Django and Django REST Framework with real-time and asynchronous components. Channels supports communication workflows, while Celery and Redis provide background-processing infrastructure.

The experimental AI stack includes locally oriented components such as Ollama, Whisper.cpp and Piper for language processing, speech recognition and speech generation.

OUTCOME

The project remains under active development and experimentation.

It currently demonstrates the architecture required for AI-assisted reception and voice automation. Planned capabilities should remain clearly separated from features that are already implemented.