Audio AI Receptionist Development
Budget: $250 – $750 USD
I want to roll out an AI receptionist whose single role is to speak with customers and give them clear, accurate information about our products and services. A caller (or site visitor) asks a question, the system turns that speech into text, pulls the correct answer from a knowledge base I will supply, and then replies in smooth, human-sounding audio—no appointment booking, no generic FAQ chatter, just concise product guidance.
Here is the scope I have in mind:
• Speech-to-text captures the question.
• An NLP layer (LLM, Dialogflow, Rasa, or a framework you recommend) matches the query to the right answer.
• Text-to-speech returns the response in natural audio, ideally with a custom voice so it feels branded.
• The module plugs into either my website chat widget or a phone line; I’m flexible as long as latency stays low and audio quality is crisp.
Acceptance criteria
– Handles at least 90 % of sample product questions correctly.
– End-to-end response time under two seconds.
– Audio replies sound natural and welcoming, not robotic.
Please tell me which tools or APIs you prefer (Whisper, ElevenLabs, Azure Cognitive Services, etc.), how you’ll ingest and update the knowledge base, and share any voice-AI work you’ve done before. Source code and clear setup notes must be part of the final deliverable so I can host it myself once the project is complete.
Here is the scope I have in mind:
• Speech-to-text captures the question.
• An NLP layer (LLM, Dialogflow, Rasa, or a framework you recommend) matches the query to the right answer.
• Text-to-speech returns the response in natural audio, ideally with a custom voice so it feels branded.
• The module plugs into either my website chat widget or a phone line; I’m flexible as long as latency stays low and audio quality is crisp.
Acceptance criteria
– Handles at least 90 % of sample product questions correctly.
– End-to-end response time under two seconds.
– Audio replies sound natural and welcoming, not robotic.
Please tell me which tools or APIs you prefer (Whisper, ElevenLabs, Azure Cognitive Services, etc.), how you’ll ingest and update the knowledge base, and share any voice-AI work you’ve done before. Source code and clear setup notes must be part of the final deliverable so I can host it myself once the project is complete.