金融科技 AI MS Chatbot

Multilingual support
Voice input
Instant recognition of pictures
Multi-platform support

Inspiration

Artificial Intelligence (AI) technology matures, according to a number of research institutes conducted a survey, AI chatbot plays an essential role in fintech. However, most bank’s chatbot is not consummate. The reason is that they cannot support the user's basic requirements. Take the Hang Seng Bank robot as an example, they only support two languages and one platform, they also do not support voice input. However, these requirements can make user feel more useful and convenience. Therefore, we hope we can use Microsoft Azure the solve the problem.

What it does

Our chatbot have Multi-platform support, multi-language support, Voice input, Instant recognition of pictures and Natural language processing. User hope the chatbot can support their native language so them can use the chatbot without learn a different language. For the company, it means that can contact more people and make more money. Using the Translator Text API, our chatbot can support more than 60 languages. User also hope they don’t need to input text but can direct ‘talk’ to the chatbot.

How we built it

First of all, we are using Azure Bot Service for our base platform. Next, we use the Microsoft Bot Framework to connect messaging channels. We also use more than one API service to build our chatbot, such as Bing API, Translator Text API and Computer Vision API. Moreover, we are using Translator Text API to make multi-language support. Translator Text API translate all revenue languages into English and send them to Azure Bot Service, then translate all the results into the language they want, and send it to the user. On the other side, we also use Voice API to Convert voice to text, so user can input voice. And we also use Voice API to return a mp3 to user, therefore, they do not need to read the text but can listen to voice. Computer Vision API

Challenges we ran into

Although Translator Text API can support 60 languages, but there are only more than then language can, so we can’t cover all user in different place.

Accomplishments that we're proud of

We are proud of using the cloud service since it provides serviceless. We are proud of the multi-language voice input. Also, we support the image input and multi-platform

What we learned

In this project, we learn many function in Azure such as : Translator Text API, Computer Vision API and some the most advances technology. We using the Azure cloud service to solve many problem such as: we can in put the voice , we can translate the all revenue languages.

What's next for 金融科技 AI MS Chatbot

In the future, we hope our project will have analysis big data to analysis the use cause by the user to know the user habit, so the bank can learn from some data, and improve their plan. Because of the data, bank will know what the public like or not like, so bank can against this data to improve the user experience.