This project is part of the Artificial Intelligence course at FIAP - Online 2025. This repository contains the solution for the Global Solution "Future at Work".
Proximity AI is an AI-powered recruitment matching platform that uses semantic intelligence to connect candidates and recruiters through simple PDF uploads. By eliminating traditional forms and keyword-based searches, Proximity reduces friction in the recruitment process, making talent matching more human, efficient, and accessible.
This project addresses the Global Solution 2025.2 theme: "Future at Work", responding to the challenge: "How can technology make work more human, inclusive, and sustainable in the future?"
Proximity AI focuses on the thematic axis: "Ethical Data-Driven Recruitment and Inclusion", creating a platform that:
- Humanizes the recruitment process by eliminating repetitive form-filling and manual data entry
- Promotes inclusion through semantic matching that understands context and intent, not just keywords
- Saves time for both recruiters and candidates, allowing them to focus on meaningful human connections
- Ensures privacy with full user control over data deletion and transparency in data processing
This solution integrates knowledge from multiple disciplines:
- AI for RPA: Document understanding and automated data extraction from PDFs
- Front End and Mobile Development: Responsive web application with mobile-first design (PWA-ready)
- Generative AI: Content generation, moderation, and intelligent resume matching using Gemini models
- AI Governance: Ethical AI practices, content moderation, and transparent data handling
- Natural Language Processing: Semantic embeddings and natural language understanding for intelligent matching
- Computer Vision: OCR and document processing for PDF extraction
- Cloud Computing: Full-stack deployment on Cloudflare Workers infrastructure
- Data Science: Vector databases, embeddings, and semantic search using Qdrant and Cloudflare Vectorize
- Frontend: React with TanStack Router, shadcn/ui components, and Tailwind CSS
- Backend: Cloudflare Workers with D1 (SQLite), R2 (blob storage), and Vectorize (embeddings)
- AI: Google Gemini for text generation, embeddings, and document extraction
- Authentication: BetterAuth with GitHub and LinkedIn OAuth
- Vector Database: Qdrant for semantic search and similarity matching
To run this project locally, follow the steps below:
git clone https://github.com/luisfuturist/gs-faw.git
cd gs-faw
pnpm install
pnpm db:migrate
pnpm run up # Start the Qdrant container
pnpm vdb:migrate # Create the collection in Qdrant
pnpm start- Caio Castro (RM559766)
- Felipe Soares (RM560151)
- Fernando Segregio (RM559582)
- Luis Emidio (RM559976)
- Wellington Brito (RM552157)
- Tutor: Leonardo Ruiz Orabona
- Coordenador: André Godoi