---
title: AI Recruiting Agent
emoji: 🤖
colorFrom: indigo
colorTo: blue
sdk: gradio
python_version: "3.10"
app_file: app.py
fullWidth: true
header: default
short_description: Bias-aware RAG AI recruiting agent with hallucination checks
suggested_hardware: cpu-upgrade
pinned: true

tags:
  - gradio
  - langchain
  - chroma
  - rag
  - recruiting
  - hr-tech
  - responsible-ai
  - fairness
  - bias-mitigation
  - llm
  - vector-database
  - ai-agent
  - nlp
  - demo
  - explainable-ai
  - enterprise-ai

models:
  - openai/gpt-oss-120b

thumbnail: https://huggingface.co/spaces/19arjun89/AI_Recruiting_Agent/raw/main/thumbnail.png

disable_embedding: false
startup_duration_timeout: 45m
---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

# 🧠 AI Recruiting Agent

A bias-aware, verification-driven recruiting assistant built with Gradio, LangChain, and vector search.  
This Space is designed as a **decision-support tool** to help recruiters assess candidate fit and generate outreach emails — while embedding safeguards for fairness and transparency.

---

## 🚀 What This Space Does

### 1) Candidate Assessment (Recruiter View)
- Upload **company culture documents**
- Upload **resumes in bulk**
- Paste a **job description**
- The system evaluates each candidate across:
  - Technical skills match
  - Culture fit
  - A final hiring recommendation
  - Claim verification against source inputs
  - A structured **bias audit**

### 2) Cold Email Generator (Candidate View)
- Upload a single resume
- Paste a job description
- Generates a **tailored professional cold email** for outreach

---

## 🛡️ Responsible AI Safeguards

This prototype embeds multiple layers of protection against hallucination and bias.

### 🔹 Input Anonymization
Resumes are sanitized before embedding and analysis:
- Emails, phone numbers, URLs, addresses, and explicit demographic fields are redacted
- Likely name headers are masked

This reduces demographic leakage into both vector search and LLM reasoning.

---

### 🔹 Fact Verification
All skills and culture analyses are checked against:
- Resume content
- Job description
- Culture documents

Unsupported claims are flagged and can trigger a self-correction routine.

---

### 🔹 Bias Audit Chain
For each candidate, the assistant runs a dedicated **bias audit** prompt that inspects:

- Over-reliance on education pedigree or past employers
- Penalization of nontraditional career paths
- Subjective or exclusionary language in cultural fit
- Reasoning not grounded in the job description or culture docs

The output includes:
- **Bias Indicators**
- A **Transparency Note** for recruiter review

These signals do **not** disqualify candidates automatically — they flag where human judgment is critical.

---

## 🧪 Try It Out

1. Upload sample culture documents (PDF or text)
2. Upload resumes (PDF or text)
3. Paste a job description
4. Click **Analyze Candidates**

Or:

- Upload a single resume  
- Paste a job description  
- Click **Generate Cold Email**

---

## ⚠️ Human-in-the-Loop Disclaimer

This tool is intended as **decision support only**.  
It does not replace recruiter judgment, legal review, or organizational hiring policies.

Final hiring decisions must always be made by humans.

---

## 🧱 Technical Stack

- **Gradio** for UI
- **LangChain** for LLM orchestration
- **Chroma** for vector storage
- **ChatGroq** for LLM inference
- **Hugging Face Embeddings** for semantic search

---

## 📣 Feedback Welcome

This is an early prototype. I’d love feedback on:

- Bias mitigation approaches
- Evaluation transparency
- UX improvements
- Failure modes or edge cases
- Responsible AI design patterns

Please share feedback via the **Community** tab on this Space.

---

*Built by Arjun Singh*