Pitch Deck

Link: https://youtu.be/zNXAqhlvM0A


Having gone through the painful process of searching for my first 'proper' job several years ago I found the process painful and very time-consuming. What I needed was a way to quickly find jobs that suited my skill set and advice on how I could get the jobs I really wanted.

What it does

Skill Up saves time by analyzing job descriptions and showing you your key strengths for a role and areas for development. Helping you to Skill Up and get your perfect job.

How we built it

Skill Up takes the 'Skills' & 'Behaviours' that users input into the app and uploads them to Google Firestore. Users are then able to browse for jobs that they think they might be suitable for. Once the 'Analyse Job Description' button is pressed the HTML from the website is extracted and saved back to Firebase. We added the BERT Q&A Model into the app using Tensor Flow. An analysis function is called which retrieves the user Skills, Behaviour & Job description text from Firebase. This is then used to work out if any of the skills/behaviors are present, using fuzzy matching.

Challenges we ran into

Due to the complexity and tight time restriction, we ran out of time to properly implement the model. If we were given more time I feel confident that we would have been able to get the ML side of the app working. This was our biggest challenge

Accomplishments that we're proud of

The three of us are really happy with the way the UI/UX of the app turned out. We think it's clean and easy to use and isn't far off looking like a production application.

What we learned

Spend less time on the front end and more time on ML development. We underestimated how long this would take by quite a lot. We learned how to go about implementing ML into native iOS app using pre-trained models which super interesting.

Built With

Share this project: