Voice Product Support AI
Budget: ₹12,500 – ₹37,500 INR
I need an AI-powered voice assistant built and ready to deploy within the next five days. Its sole purpose is customer support: when a user speaks, the assistant should recognise the inquiry, pull the right product details from my existing database or API, and respond naturally in real time.
Key requirements
• Voice-first interaction: speech-to-text, intent detection, and text-to-speech all running with minimal latency.
• Product information only: the assistant does not handle FAQs outside that domain, schedule appointments, or manage email—just accurate product answers.
• Easy hand-off: if the question falls outside its scope, a smooth transfer to my human support line is essential.
I’m happy for you to choose the stack—Dialogflow, Rasa, Azure Cognitive Services, Amazon Lex, or a custom PyTorch/NLP pipeline—so long as the final solution reaches production-ready performance on both web and mobile channels. Provide the full source code, setup documentation, and a short video demo proving the assistant works end-to-end.
Delivery deadline is five days from project start, with daily progress check-ins so I can test early builds and keep us on track.
Key requirements
• Voice-first interaction: speech-to-text, intent detection, and text-to-speech all running with minimal latency.
• Product information only: the assistant does not handle FAQs outside that domain, schedule appointments, or manage email—just accurate product answers.
• Easy hand-off: if the question falls outside its scope, a smooth transfer to my human support line is essential.
I’m happy for you to choose the stack—Dialogflow, Rasa, Azure Cognitive Services, Amazon Lex, or a custom PyTorch/NLP pipeline—so long as the final solution reaches production-ready performance on both web and mobile channels. Provide the full source code, setup documentation, and a short video demo proving the assistant works end-to-end.
Delivery deadline is five days from project start, with daily progress check-ins so I can test early builds and keep us on track.