Demonstration: Part 1
Demonstration: Part 2
Demonstration: Part 3
Good Candidate: Post Processing Analysis
Bad Candidate: Post Processing Analysis
The custom areas to grant points at.
The Origin Story
As a student, I often hear how nervous my peers get about job interviews. Company recruiters also wish that interviewing was more time efficient and less emotionally draining. If neither parties like how job interviews are done, then shouldn't we change the whole process?
Recruiting technical talent that can communicate is ideal. Many potential employees are turned down without knowing what they need to improve. Some people gain the opportunity to interview by using "white-text" to cheat resume algorithms while these algorithms also overlook talented individuals. In addition, it is very hard for employers/recruiters to make an accurate judgment on someone after only the first meeting.
Employers/recruiters have little to no assistance when it comes to determining how well an interviewee fits the job. Furthermore, interviewing many people in one day will hinder recruiters' ability to focus on and to assess their entire applicant pool.
Our InterviewBot seeks to revolutionize the interviewing process so it will be fun for both the interviewer and the interviewee. We aim to eliminate fragmented interviews and to provide a truly comprehensive, authentic interview process.
•Videochat and screensharing eliminates the need for in-person interviews thereby minimizing travel delays, can be used to interview potential employees independent of their global location while ensuring that they're authentic.
•Speech-to-text, text-analysis and our weighted score table accurately determine how well the interviewee fits the job and compromises for the lack of an in-person interview. This also eliminates the need for multiple rounds of interviews thereby streamlining the process
•Gamification makes the interview a fun process and less stressful for potential employees
How we built it
Challenges we ran into
At first, it was difficult to change in the interviewing process in a way that would benefit both the interviewer and the interviewee. The screensharing and videochatting consisted of Open Broadcast Software ,the Screenleap program. We stripped the audio for language processing in IBM, but it took a bit of time to get all the components to work together. The slow internet connection and the lack of food at the hackathon was also an obstacle. Luckily, the teamwork, communication skills and patience of our group enabled us to create a quality project.
Accomplishments that we're proud of
Our group took on a very difficult problem and was able to produce a suitable project in order to solve it. Getting a comprehensive streaming service was accomplished inside the timeline by re-inventing the wheel as little as possible. We realized that Twitch.tv, Open Broadcast Software, Screen Leap, and some front-end design along with some creative acting could make our product shine. We took a lot of time researching and have more content than 3 minutes can reflect however, we are confident that this is a realistic startup company idea focused on B2B for the large corporate entities like Google and IBM.
What we learned
In order to fully improve the interviewing process, we will need both companies and workers to acknowledge that the traditional in-person interview's time era is over.
What's next for InterviewBot
We need clients with custom requests to process data on a live bias sample. Ideally, our early adopter would provide this type of insight so that we can further fine-tune our process and ensure our analytics stays accurate and reliable. The next iteration of this product would include the facial expression recognition to further optimize the non-verbal queue of an interview for the employer to decide how heavily they'd weigh in (in case it's a tech-service industry).