The complete project is to develop a mobile application that provides the functionality of a Chatbot and a Q&A dialogue session. The final application will use a decision tree to develop the best response to the user’s input however this initial stage is to develop a test platform. The immediate objective is not to develop a finished application but rather the framework upon which to develop the intelligence. I am a student experimenting with AI and this project will provide the scaffolding for this activity.
I have decided on the following development tools/cloud model but am open to recommendations for alternative approaches as it has been some years since I’ve been in the development industry. My development environment is Visual Studio 2015 Community running on Windows 10 with access to a limited set of Azure services. All development will need to fit back into this environment so that I can continue to develop.
The end users will access a chatbot service via a mobile app. Mobile app development should be in [url removed, login to view] as I wish to use a single tool to develop for iOS, Android, and Windows. The mobile interface to the end user does not require complex mobile interactions so [url removed, login to view] should be suitable for all platforms.
The mobile app will interface with an Azure hosted Mobile/Web App. This app will receive input from the mobile device, manipulate the input and return a response to the mobile device.
The system requires a username/password to authenticate the user and/or create a new account. This should use the Azure Authenticator App service. If possible, after the initial authentication the mobile device should continue to be recognised.
The mobile app should allow a user to enter a sentence as speech and then to provide an audio response. Many similar chatbot agents use a ‘press to talk’ button to more easily determine the start and end of the user’s spoken dialog however other methods can be considered.
As the user speaks and the system responds a dialogue frame on the mobile device will show the entire transcribed history of the interaction. An example follows:
User: my cat died.
ChatBot: that sucks. How old was it?
User: she was 17 years.
ChatBot. Wow, that's old for a cat!
The user entered audio is to be sent from the mobile device via the Internet to Azure Cognitive Services (Speech to text) API, converted to text and then displayed back to the user.
[url removed, login to view]
This text is to be stored in in an Azure database structure (SQL or MySQL) along with user details and a timestamp.
The text is then to be passed to a Python (V3.4+) program for processing.
The python program needs to process the text using 6 different processing methods and the responses reviewed and prioritised to select a best response.
For this project, a random selection can be used to determine the best response and the program can respond with (‘Response X’ + ‘is the way to go’).
The best response will then be sent to the Azure Cognitive Services (Text to speech) API and the audio and transcription presented back to the user.
The database should store user profiles and all user conversations by interaction.
All interactions are to be stored to an unencrypted database however in the future there may be a need for this to become encrypted.
There may be a need to access this same service via a web portal in the future.
The proposals I've received are very varied, as are the skill sets of the freelancers and so I wish to be clear about what is required.
- mobile chatbot application written using Xamarin for iOS, Android, and Windows, that interfaces with a web app running on Azure.
- mobile app session commences with user authentication using Azure and Microsoft Authenticator service
- there is a need to create a simple Azure MS-SQL database to store the User data and conversation interactions.
- a web app running on Azure that converses with the mobile app. This app manages the voice to text and text to voice translation via Azure Cognitive speech tools.
- the web app and Xamarin are likely to be written in C#
- a stub Python program that provides an API that receives the user's text. This program processes the text into a simple response and then transfers it back to the web app. The web app generates a text to voice response and both the text and audio are sent back to the mobile application.
I will provide access to my Azure account subscription to allow access to the Azure services for authentication, database services, and cognitive services such as voice-to-text and text-to-voice.
14 фрилансеров(-а) в среднем готовы выполнить эту работу за $805
i have gone through your requirement we done similar kind of job before looking forward your earliest Reply on this for a project discussion Awaiting for your earliest reply