Inspiration

We were inspired by the countless stories of underpaid and exploited workers, particularly migrants and young people in rural areas, who often lack awareness of their rights or are too afraid to speak out.

We asked ourselves: What if there were a digital tool that could safely and privately guide them—like a virtual union rep?

This question sparked the creation of KnowYourWorth, a platform designed to empower workers through clear information, decision-making support, and AI-enhanced tools, even for those with limited legal or digital literacy.


What it Does

KnowYourWorth is a web application that helps workers evaluate their job conditions and take informed action. Its main features include:

  • Contract & Salary Check:
    Workers can input their contract type, salary, and job role to compare against official national collective agreements (CCNLs) and detect potential violations.

  • Decision Quiz:
    A short, guided quiz helps users determine whether it is safe and appropriate to contact a union, and explains how to do so if needed.

  • PDF Action Plan Generator:
    In cases of potential exploitation, the app generates a downloadable document tailored to the user's situation. It outlines recommended next steps and how to take them based on the data provided.

  • AI-Powered Insights:
    Information collected from the user is refined through various AI modules to improve the accuracy and relevance of the advice offered.


How We Built It

  • Frontend: Built with Streamlit for a clean, accessible, and user-friendly interface.
  • Backend: Developed using Python and Flask, with dynamic logic and automated PDF generation.
  • Perplexity API: Used for user input enhancement, query refinement, RAG (Retrieval-Augmented Generation), and personalized advice generation.
  • LlamaIndex: Implemented for structured and efficient RAG processes.
  • OpenAI Embeddings: Used for vector search and similarity matching within our document database (CCNLs), chosen for deployment efficiency.
    > The system is also compatible with Hugging Face embeddings, allowing flexibility for future self-hosted or multilingual use cases.

All sensitive data is processed locally or anonymized before interacting with external APIs, ensuring user privacy.


Challenges We Ran Into

  • Balancing the complexity of labor laws with a simple and intuitive UX.
  • Developing a reliable CCNL comparison engine despite limited access to structured labor data.
  • Managing API rate limits while maintaining quality user experience.
  • Ensuring all content and features are safe, accurate, and actionable, especially for vulnerable populations.

Accomplishments We're Proud Of

  • Built a functional MVP in just 10 days.
  • Created a tool that delivers real value to underserved workers.
  • Integrated AI in a transparent, ethical, and socially responsible manner.
  • Designed a legally-aware UX focused on building trust with users who may be skeptical of institutions.

What We Learned

  • Legal tech solutions require not just functionality, but empathy and clarity.
  • Even powerful tools like Perplexity demand careful prompt crafting and responsible usage.
  • Designing for marginalized users means prioritizing simplicity over sophistication.

What's Next for KnowYourWorth

  • Expand support for additional sectors such as hospitality, logistics, and agriculture.
  • Introduce a Statistics Dashboard that aggregates user data by job sector and region, helping institutions identify systemic issues and prioritize interventions transparently.
  • Launch a Wage Benchmarking Tool to help users understand how their pay compares to regional and industry standards.
  • Collaborate with unions and NGOs to pilot-test the platform in real-world settings.
  • Develop features for anonymous reporting and secure data storage to support users over longer-term cases.

KnowYourWorth strives to be a digital ally for workers—helping them understand their rights, compare their conditions, and take safe, meaningful steps toward justice.

Built With

Share this project:

Updates

Submission history