bloom

Inspiration

To improve social mobility among low-income women, employment is a key consideration. There is a white space in today's landscape, where current job boards do not adequately consider soft skills and disadvantage our target audience of low-income women, and current initiatives for the target audience are not sufficient in supporting the long-term employment of these women. Thus, our project was formed.

What it does

Our project aims to connect women with less social mobility and without access to livelihood opportunities to employers based upon traits that go beyond their resume: by generating recommendations on the basis of both soft skills and competencies, we aim to better represent the true value of this target group and help employers see the value in this untapped pool of resilient and courageous women who have overcome various difficulties in their diverse journeys.

How we built it

Our technical prototype makes use of a natural language processing model built with spaCy. This model consists of a Named Entity Recogniser trained on a dataset of annotated resumes and personal statements to detect general soft sklils, and rule-based matching for specific phrases of soft skills as well as technical competencies.

This model is used in our Streamlit app, which has been deployed on Google Cloud Run.

This mirrors the portion of our product where applicants answer a set of open-ended behavioural questions to generate soft skills, as well as resume-parsing for technical competencies. The same model is also used to parse testimonies that the applicant provides. On the employer's side, we also aim to identify the technical skills needed based on the job description and responsibilities, while allowing them to manually specify the soft skills that they value based upon their company culture and expectations for the role. Employers and employees with the highest compatibility in terms of matching both soft skills and competencies, are recommended to each other.

Challenges we ran into

Understanding the user base and obstacles they face on the daily were key challenges were ran into as understanding the consumer's pain points is key to creating a comprehensive and effective solution. As such, we conducted secondary research to consider key challenges faced and what aspects would be helpful in solving their problem to improve social mobility for low-income women.

Accomplishments that we're proud of

Our key deliverables of the NLP model, Figma prototype and design, as well as pitch deck that were all done up in a matter of days are accomplishments our group is proud of, especially since we were utilising tools unfamiliar to us and with a steep learning curve.

What we learned

Not only have we improved our proficiencies in the relevant technical tools, we have also learnt the importance of diversity and teamwork in a team with varied backgrounds. Being able to build up a new idea from scratch to a stage where it is highly developed has been extremely insightful in understanding the ideation, creation and pitching process.

What's next for bloom

Next, we hope to continue to build up our NLP model to increase its accuracy in generating valuable soft skills. Furthermore, we hope to gather more user research to understand pain points and work on these aspects. Moving forward, we also hope to enable these women to take part in courses, be it relating to soft skills or competencies, in order to quickly up-skill for increasing social mobility.