Inspiration
We wanted to make make an easy way to search Craiglist for free items.
What it does
Our Backend is responsible for communicating with the Craigslist API and dialogflow's Node.js client and exposes a REST API to connect the processes.
How we built it
We built a backend with Express, powered by Node.js. It's responsible for communicating with the Craigslist API and exposing a REST API. Our backend also plugs into a script we wrote to interface with our dialogflow agent using its node.js client. Our backend is responsible for most of the logic to query free products from Craigslist.
Our dialogflow agent utilizes machine learning and natural language processing to extract the product in contention based on our provided annotations and training data. Our model uses a human-in-the-loop technique to continuously improve the quality of our machine learning model.
We developed our slack bot using Node.js as well and communicate with our backend to relay information from Craigslist.
Challenges we ran into
We had some trouble with authenticating our slack app to our localhost server. We solved this by using Ngrok, a tunneling software that created a secure tunnel to the server.
Accomplishments that we're proud of
What we learned
A lot of us had very little or no Node.js experience, so it was a nice learning experience.
What's next for Freelist
Built With
- bulma
- dialog-flow-api
- express.js
- html
- machine-learning
- node.js
Log in or sign up for Devpost to join the conversation.