57% of Donators can not find an appropriate or aligned campaign for them to donate whether it’s the project’s description, urgency, funding, or amount of money. And 78% of donors would donate money to pets. And each donor has donated more than once.
What it does
- Donors can ask (input) the questions about the donation containing what type of donation and amount of money they want to donate within the whole sentences to the chatbot and then it will answer (output) related to the specific goal of donation and number of money that donors ask.
- AI Chatbot will choose to donate randomly within the corporation pool.
- AI Chatbot have the choices to choose an item to be donated as the user wishes.
- User chooses the amount of donation and the chatbot will distribute automatically in criteria priority.
How our team built it
The chatbot is used to automate responses by using the keyword rules built right into the dashboard. It allows the AI bot to “understand” and learn from the messages it receives, take actionable data from those messages, and deliver even more accurate responses. Using Node.JS to create AI Chatbot and connect it with LINE Official Account API to be input and output device for this prototype.
- Define the goals: to increase customer satisfaction .ex. we might want to add a chatbot to the customer to the support team and let it handle the most common FAQs, so our team can focus on the more complicated cases.
- Define the use cases: Our team needs to figure this out before starting building the bot,as we need to know exactly what the bot will do and why that is important.
- Craft bot personality: We want to make sure that our users connect with the chatbot and that the conversation is engaging and representative of real human interaction.
- Map out the path: create a logic wireframe to see how a user would go from start to finish and where they might want to dive into other flows.
- Write a Script for edge cases: the sample dialog should help pinpoint the pain points and off-track problems, as will user testing via Dialogflow.
- Bot Testing/Optimising: Internal testing will give a lot of insight on how to improve AI bot. So after publishing the bot, we need to keep monitoring its performance. Monitor the conversations, collect data, create logs, analyze the data, and keep improving the bot for an even better experience.
Challenges our team ran into
At first, We decide to use Dialogue Flow because it's a chatbot tool that we are experts at, but after we research more. We found out that it was unable to calculate the amount of money since it's not suitable. We switched to Node.Js instead, plot the graph with Python, and then connect it to LINE OA API.
Accomplishments that I'm proud of
Be able to save our time and money for communicating with the customer and gain customer satisfaction. Also, be able to suggest customers donate based on the criteria priority.
What I learned
- The more data we use, the more distributed AI can suggest
- Each corporation has different needed
- We can find a prediction of AI distribution behavior after it runs for a long time
What's next for PetFund.Co
- FAQ by using AI to automatically answer.
- Increase AI-powered chatbots features as a suggestion or automate advertisement campaigns/events specific for each of the user’s interests.
- PetFund.co will apply more features implemented by AI in the platform by using AI into the Blockchain system to make funding procedures more secure and entirely open.