🚀 HirePilot — AI-Powered Candidate Ranking Platform

Redrob Hackathon 2026 Submission Transforming recruitment with AI-driven candidate discovery, ranking, explainability, and fraud detection.

---

📌 Problem Statement

Recruiters often receive thousands of applications for a single job opening, making manual screening slow, inconsistent, and inefficient.

HirePilot solves this problem by automatically analyzing candidate profiles, matching them against a Job Description, detecting suspicious profiles, and generating an explainable ranking of the most relevant candidates.

---

🎯 Solution Overview

HirePilot is an end-to-end AI recruitment intelligence platform that:

• Parses and analyzes large candidate datasets
• Matches candidates with job descriptions using semantic AI
• Scores candidates across multiple dimensions
• Detects fraudulent or manipulated profiles
• Generates explainable recommendations
• Produces recruiter-ready ranked candidate lists

The platform combines Natural Language Processing, Machine Learning, behavioral analysis, and rule-based intelligence to help recruiters identify the best talent faster.

---

✨ Key Features

🤖 AI-Powered Candidate Ranking

• Semantic Job Description matching
• Resume-to-role relevance scoring
• Multi-dimensional candidate evaluation
• Intelligent ranking engine

---

🧠 Semantic Matching Engine

Uses Sentence Transformers to understand:

• Skills relevance
• Role similarity
• Industry alignment
• Contextual candidate-job fit

Unlike keyword matching systems, HirePilot understands meaning and context.

---

📊 Multi-Factor Scoring Framework

Candidates are evaluated across:

Dimension	Weight
Skill Match	30%
Semantic Similarity	25%
Experience	20%
Behavioral Signals	15%
Education	10%

---

🛡 Fraud Detection Engine

HirePilot automatically detects:

• Keyword stuffing
• Fake experience claims
• Timeline inconsistencies
• Honeypot profiles
• Education anomalies
• Incomplete profiles
• Role-history contradictions

Suspicious candidates receive score penalties.

---

💡 Explainable AI

Every ranked candidate includes:

• Score breakdown
• Matching strengths
• Ranking justification
• Recruiter-friendly reasoning

This ensures transparency and trust in AI decisions.

---

📈 Recruiter Analytics Dashboard

Visual insights including:

• Candidate ranking distribution
• Score analytics
• Skill coverage analysis
• Candidate quality metrics

---

🏗 System Architecture

Candidate Dataset
        │
        ▼
 Resume Parsing Layer
        │
        ▼
 Feature Extraction Engine
        │
        ▼
 Semantic Matching Engine
        │
        ▼
 Multi-Factor Scoring
        │
        ▼
 Fraud Detection Layer
        │
        ▼
 Ranking Engine
        │
        ▼
 Explainability Module
        │
        ▼
 Top Candidate Recommendations

---

📂 Project Structure

backend/
├── api/
├── parser/
├── scoring/
├── ranking/
├── behavioral/
├── fraud_detection/
├── explainability/
├── validation/
├── config.py
├── pipeline.py
├── main.py
└── requirements.txt

frontend/
├── src/
├── public/
├── package.json
└── vite.config.ts

data/
├── candidate_schema.json
├── candidates.jsonl
└── job_description.txt

scripts/
├── generate_mock_data.py

tests/
├── test_parsers.py
├── test_scoring.py
├── test_ranking.py
└── test_validation.py

---

⚙ Technology Stack

Backend

• Python 3.11
• FastAPI
• Uvicorn
• NumPy
• Pandas
• Scikit-Learn
• Sentence Transformers

Frontend

• React 18
• TypeScript
• Vite
• Tailwind CSS
• Zustand
• Axios
• Recharts

AI & NLP

• Sentence Transformers
• Cosine Similarity Matching
• Feature Engineering
• Explainability Framework

---

🚀 Getting Started

1️⃣ Clone Repository

git clone <repository-url>
cd HirePilot

---

2️⃣ Backend Setup

cd backend
pip install -r requirements.txt

---

3️⃣ Frontend Setup

cd frontend
npm install

---

4️⃣ Start Backend

cd backend
uvicorn main:app --reload --port 8000

Backend:

http://localhost:8000

API Docs:

http://localhost:8000/api/docs

---

5️⃣ Start Frontend

cd frontend
npm run dev

Frontend:

http://localhost:5173

---

📡 API Endpoints

Method	Endpoint	Description
POST	/api/upload/candidates	Upload candidate dataset
POST	/api/upload/jd	Upload Job Description
POST	/api/upload/jd-text	Submit JD text
POST	/api/process	Start ranking process
GET	/api/status	Processing status
GET	/api/candidates	Ranked candidates
GET	/api/candidates/{id}	Candidate details
GET	/api/top100	Top-ranked candidates
GET	/api/export	Export rankings
GET	/api/validate	Validate submission
GET	/api/analytics	Analytics dashboard

---

🧪 Testing

Run all tests:

pytest tests/ -v

---

🎯 Hackathon Highlights

✅ AI-Powered Recruitment Intelligence

✅ Semantic Resume Matching

✅ Explainable Candidate Ranking

✅ Fraud Detection System

✅ FastAPI + React Full Stack Architecture

✅ Recruiter Analytics Dashboard

✅ Automated Validation Pipeline

✅ Scalable Architecture for Large Candidate Pools

---

🔮 Future Roadmap

• LLM-powered Interview Question Generation
• Resume Summarization
• ATS Integration
• Recruiter Feedback Learning Loop
• Multi-language Resume Support
• Advanced Candidate Recommendations
• Real-Time Hiring Insights

---

👥 Team

🚀 HirePilot — Redrob Hackathon Team

Team Member
Swagatam Behera
Ashwath Raj
Sahil Ranjan
Kaushik Katual

---

📜 License

This project is licensed under the MIT License.

---

⭐ Built for Redrob Hackathon 2026

Empowering recruiters with AI-driven hiring intelligence.