Hiring Agent: An AI Agent for Resume Evaluation and Scoring

Introduction

Hiring Agent is a powerful open-source AI tool developed by InterviewStreet, designed to streamline and enhance the resume evaluation process. It automates the extraction of structured data from PDF resumes, enriches this data with valuable GitHub profile and repository signals, and then generates an objective, explainable evaluation. This system aims to provide a consistent and fair scoring mechanism for candidate resumes.

The agent's core functionality involves parsing resume PDFs into a markdown-like format, extracting key information using large language models (LLMs), and augmenting this data with insights from a candidate's GitHub activity. The final output includes category scores, supporting evidence, bonus points, and deductions, offering a comprehensive overview of a candidate's profile. It offers flexibility by supporting both local LLM setups with Ollama and cloud-based solutions like Google Gemini.

Installation

Getting started with Hiring Agent is straightforward. Follow these steps to set up the project on your local machine:

Prerequisites

• Python 3.11+: The repository is configured for Python 3.11.13.
• One LLM backend: Choose either Ollama for local models or Google Gemini if you have an API key.
  • Ollama: Install from the official site, then run ollama serve.
  • Google Gemini: Obtain an API key from AI Studio.

Quick setup with pip

$ git clone https://github.com/interviewstreet/hiring-agent.git
$ cd hiring-agent

$ python -m venv .venv
# Linux or macOS
$ source .venv/bin/activate
# Windows
# .venv\Scripts\activate

$ pip install -r requirements.txt

Ollama Models

If you're using Ollama, pull the desired model. For example:

$ ollama pull gemma3:4b

Configuration

Copy the example environment file and configure your LLM provider and model:

$ cp .env.example .env

Edit the .env file to set LLM_PROVIDER (e.g., ollama or gemini), DEFAULT_MODEL (e.g., gemma3:4b or gemini-2.5-pro), and GEMINI_API_KEY if using Gemini. Optionally, set GITHUB_TOKEN for improved GitHub API rate limits.

Examples

To score a resume end-to-end, simply provide the path to a PDF file using the command-line interface:

$ python score.py /path/to/resume.pdf

When you run this command, the following actions occur:

1. If DEVELOPMENT_MODE is enabled in config.py, the PDF extraction result is cached.
2. If a GitHub profile is identified in the resume, repositories are fetched and cached.
3. The evaluator generates a detailed report, which is printed to standard output. In development mode, a CSV row is also appended to resume_evaluations.csv.

Why Use It

Hiring Agent offers several compelling advantages for anyone involved in the hiring process:

• Objective and Explainable Evaluations: It provides a structured scoring system with clear evidence, reducing bias and increasing transparency in resume assessments.
• GitHub Enrichment: By integrating GitHub profile and repository data, it offers a deeper, more holistic view of a candidate's technical contributions and project experience.
• Flexible LLM Support: The ability to use both local LLMs (Ollama) and cloud-based LLMs (Google Gemini) provides flexibility in deployment and cost management.
• Modular Architecture: Its well-defined modules for PDF parsing, data extraction, GitHub integration, and evaluation make it extensible and easy to maintain.
• Open-Source: Being open-source, it allows for community contributions, customization, and continuous improvement, adapting to evolving hiring needs.

Links

• GitHub Repository: https://github.com/interviewstreet/hiring-agent