Inspiration
We were inspired by our own experience and by the reality faced by many people in the Latin community who want to enter the tech industry but struggle to understand where to start. Job descriptions are often overwhelming, written in complex language, and filled with unnecessary requirements. For Spanish-speaking users, this challenge is even greater due to language barriers and the lack of accessible, localized guidance.
We wanted to create a tool that transforms confusion into clarity.
What it does
SkillBridge is a web application focused on tech-related jobs that analyzes job postings to extract, rank, and visualize the most relevant skills required by the market. Instead of reading long job descriptions, users can quickly see which skills matter most, how often they are requested, and their overall importance.
The platform also provides instant learning suggestions to help users begin building their skills and portfolios. The application is presented in Spanish to better support and empower the Latin community.
How we built it
- Used Python to scrape job postings from multiple job websites, focusing on job titles and requirements.
- Avoided platforms with strict scraping restrictions and adapted our data sources accordingly.
- Processed job descriptions using rule-based heuristics and a predefined skills dictionary to extract relevant technical skills while filtering out filler words.
- Stored the processed data in Supabase (free tier) as our database.
- Built the frontend using TypeScript, HTML, and CSS, starting from a mock created with Lovable and later refined manually.
- Connected the frontend to Supabase to display structured skill data and redirect users to learning platforms (currently Udemy).
Challenges we ran into
- Web scraping limitations on major job platforms such as LinkedIn.
- Extracting meaningful skills from unstructured text without relying on costly Large Language Models.
- Managing query limitations imposed by Supabase’s free tier.
- Refining the UI due to missing or non-functional components in the initial mock.
Accomplishments that we're proud of
- Built an end-to-end pipeline from job data collection to skill visualization.
- Successfully extracted relevant technical skills from real-world job postings.
- Created an accessible, Spanish-language interface focused on the Latin tech community.
- Delivered a functional prototype without requiring user login.
What we learned
- Real-world job data is messy and requires careful filtering and validation.
- Rule-based systems can still be effective when LLMs are not viable.
- Accessibility and language inclusion play a major role in user adoption.
- Building with limited resources encourages creative and efficient solutions.
What's next for SkillBridge
- Integrate official job platform APIs for real-time data.
- Expand skill extraction to include experience level and years required.
- Support multiple learning platforms beyond Udemy.
- Add personalization features and user profiles.
- Scale to support more tech roles and regions across Latin America.
Log in or sign up for Devpost to join the conversation.