Built for the Redrob Intelligent Candidate Discovery & Ranking Challenge
Finding the right candidate through recruiter-inspired reasoning rather than simple keyword matching.
Traditional Applicant Tracking Systems (ATS) rely heavily on keyword matching, often overlooking genuinely qualified candidates.
Ekalvya is a recruiter-inspired candidate ranking platform that evaluates candidates using multiple signals such as career evidence, career alignment, behavioral signals, technical relevance, and professional experience to identify the most suitable candidates.
- Intelligent candidate ranking
- Career evidence analysis
- Career alignment scoring
- Recruiter behavior signals
- Explainable recruiter reasoning
- Interactive Streamlit dashboard
- Top-100 candidate generation
flowchart LR
A[Candidate Dataset]
-->B[Feature Engineering]
B-->C[Career Evidence]
B-->D[Keyword Relevance]
B-->E[Career Alignment]
B-->F[Behavior Signals]
B-->G[Experience]
C-->H[Ranking Engine]
D-->H
E-->H
F-->H
G-->H
H-->I[Top 100 Candidates]
I-->J[Recruiter Reasoning]
J-->K[Streamlit Dashboard]
ekalvya/
├── data/
├── notebooks/
├── outputs/
├── src/
├── app.py
├── requirements.txt
└── README.md
- Python
- Pandas
- NumPy
- Scikit-Learn
- Sentence Transformers
- Streamlit
- Matplotlib
- Git & GitHub
git clone https://github.com/Keertana-N/Ekalvya.git
cd Ekalvya
pip install -r requirements.txt
python src/ranking.py
streamlit run app.py- Processed 100,000 candidate profiles
- Generated Top 100 ranked candidates
- Reduced keyword-stuffed false positives
- Explainable recruiter reasoning
- Interactive recruiter dashboard
- Semantic reranking
- Resume PDF parsing
- Vector database integration
- Resume upload
- LLM-powered recruiter summaries
⭐ Built for the Redrob Intelligent Candidate Discovery & Ranking Challenge