Lilac ChatBot

Inspiration

Many students often spend a lot of time doing online shopping. However, because each store has its own website, it takes a lot of time to consider all options and outside factors such as cost and color. This inconvenience inspired us to find a way to make this kind of shopping engaging and easy - by using a chatbox, Lilac!

What it does

Lilac is an online chatbox service that makes online shopping quick and convenient. We search a variety of retailers for the products that fit best within your desired price range and constraints. There's no need to hunt through dozens of websites individually to find the perfect article of clothing anymore; we do all the hard work for you!

How I built it

We used Dialogflow to build the framework and basic functionalities for our Lilac chatbot. We created multiple entities that classified, stored, and trained data based on expected user input. Some of the entities we created/used were Type of Clothing, Gender/age, currency, and color. Next, we used these entities to train our intents to respond to calls based on detected keyworks. For example, our general clothing intent detected and classified a type of clothing, gender, color, and price parameters after prompting the user to input details about the type of product they are looking for. Then, we sent a post request to a database on google cloud, which would find objects from the different websites that were related to the inputted keywords. These objects were then sent back to the chatbot and displayed to the user.

Challenges we ran into

Learning new frameworks Scraping from high-security retail websites Cutting down on runtime

Accomplishments that we’re proud of

We programmed a csv generator function to quickly create entities for our chat bot We brought the runtime for scraping data from websites down from a few minutes to under 5 seconds

What we learned

How to utilize dialogflow, google cloud services, bootstrap, and kommunicate Sending post requests from dialogflow How to scrape websites and filter information How AI chatboxes work with training data to identify next responses

What's next for Lilac ChatBot

Lilac still has room for improvement! By caching the data, Lilac’s runtime can be cut down, providing faster results. Additionally, we hope to add more retailers to out arsenal in the future. Perhaps out users can soon buy the product directly from our website, rather than having to go to the website of the individual retailers to complete the transaction.