AI Resume Matcher using Endee Vector Database

Overview

AI Resume Matcher is a full-stack semantic resume screening platform that helps recruiters identify the best candidates for a job description using AI-powered vector similarity search.

Built on top of the Endee vector database, the application converts resumes and job descriptions into embeddings, stores them in Endee, and retrieves the most semantically relevant resumes based on similarity search.

---

Features

• Upload multiple resumes (PDF/DOC/TXT)
• Enter job descriptions for candidate matching
• AI-powered semantic similarity matching
• Ranked candidate results with match percentage
• Modern recruiter dashboard UI
• Endee vector database integration
• Embedding-based retrieval system

---

Tech Stack

Frontend

• React (Vite)
• Tailwind CSS
• Axios

Backend

• Spring Boot
• Java
• REST APIs

AI / ML

• Embedding Generation Service
• Semantic Similarity Search

Vector Database

• Endee

---

System Architecture

Frontend (React)
      ↓
Spring Boot Backend
      ↓
Embedding Generation Service
      ↓
Endee Vector Database
      ↓
Similarity Search Results

---

How Endee Is Used

Endee acts as the vector database layer for semantic retrieval.

Workflow

1. Resume text is extracted from uploaded files
2. Text is converted into embeddings
3. Embeddings are stored in Endee with metadata
4. Job description is embedded into vector form
5. Endee performs similarity search
6. Top matching resumes are returned with scores

---

Why Vector Database?

Traditional keyword search cannot understand semantic meaning.

Using a vector database enables:

• Semantic understanding of resume/job similarity
• Better candidate ranking
• Context-aware matching beyond exact keywords

---

Project Structure

endee/
├── frontend/        # React frontend
├── backend/         # Spring Boot backend
├── docs/            # Screenshots / architecture diagrams
├── src/             # Original Endee source code
└── README.md

---

Setup Instructions

1. Clone Repository

git clone https://github.com/YOUR_USERNAME/endee.git
cd endee

---

2. Start Endee Server

docker run --ulimit nofile=100000:100000 -p 8090:8080 -v %cd%/endee-data:/data --name endee-server --restart unless-stopped endeeio/endee-server:latest

---

3. Run Backend

cd backend
mvn spring-boot:run

---

4. Run Frontend

cd frontend
npm install
npm run dev

---

API Endpoints

Upload Resume

POST /api/resumes/upload

---

Match Job Description

POST /api/match

---

Get All Resumes

GET /api/resumes

Future Improvements

• Resume parsing with advanced NLP
• Skill extraction and tagging
• Recruiter authentication system
• Job posting management
• Candidate feedback/analytics dashboard

---

Why This Project Demonstrates AI/ML + Endee Usage

This project demonstrates practical AI/ML implementation by:

• Using embeddings for semantic understanding
• Leveraging vector similarity search
• Integrating Endee as production vector database
• Building a real-world AI-powered recruitment workflow

---

Submission Notes

This project was built as part of an AI/ML assignment requiring:

• Usage of Endee vector database
• Practical AI application implementation
• Forked Endee repository as project base

---

Author

Mohamed Thousif

GitHub: https://github.com/MohamedThousif2005