InterChat

• 

  Sign-up/Login Page

• 

  Main view for candidates

• 

  Create job positing

• 

  Application report (analysis with Wit.ai)

• 

  Interview process for candidates

Inspiration

Picture this: you’re a recruiter who just created a job posting for your company. Within a few days, you receive hundreds of applications. As you’re sorting through them, you see that many of the applications don’t even meet the job requirements! You become exhausted as you read and begin to zone out. You have to set up the next round of interviews within a short period of time.

What if we could prevent this whole situation? What if there was a more efficient way to get through the first round of applicants of a job?

Introducing InterChat!

InterChat is a web application that allows recruiters to create job listings and receive a ranked list of the people who applied. The information that the recruiter is looking for will be abstracted from the responses through Wit.ai.

What it does

From the business/recruiter’s end, simply create a job listing. Enter all the criteria that the position are looking for, including location, years of experience, degree, skills, and more. Once completed your posting will be visible to all candidates using InterChat.

From the candidate’s end, they will see the job listing and apply to it. They will answer a series of questions in a chatbot. Their responses will help us determine if they have the criteria that the job posting is searching for. The information that the recruiter needs will be abstracted from the responses through Wit.ai. (e.g, how many years of experience a person has, or what degree he/she holds, etc). Moreover, the sentiment of the responses, being positive, neutral, or negative will also be obtained. These entities/traits will be sorted based on who is the most enthusiastic and qualified for the position.

How I built it

The back-end was built using Node.js with MongoDB as the database. Wit.ai was used to create utterances that would help us abstract key information (i.e, entities) we are looking for within the candidates' responses.

The front-end was built as a React application. Moreover, Ant Design, a React UI library was used to enhance the overall user experience.

Challenges I ran into

While we ran into multiple challenges, the most significant challenge was setting up the Wit.ai API connection to the application. This process was challenging due to the fact that we were unfamiliar with the documentation. However, having samples within the documentation allowed us to learn how to work with the API and successfully implement it into InterChat.

Accomplishments that I'm proud of

We are proud that the overall user experience functions smoothly. Moreover, we are glad that we were successfully able to achieve our goal of using Wit.ai to form rankings of the candidates' answers.

What I learned

We learned a lot about using the various technologies that were needed to create this project. Specifically, it was our first time using Wit.ai and it was a great experience as it helped us improve our understanding of how Machine Learning works.

What's next for InterChat

While there are plenty of aspects of InterChat that we could consider adding/improving, we would like to add an analytics page for the recruiter so he/she could see different data regarding the applicants (e.g, graphs that show the average number of years every applicant/candidate has, the most common degree, etc.).

Built With

• ant.design
• javascript
• mongodb
• node.js
• react
• wit.ai