Resume Ingestor Azure Function

A serverless Azure Function that extracts text from PDF resumes, processes them using Azure OpenAI for intelligent data extraction, and stores structured candidate information in Azure Cosmos DB for text-based search and retrieval.

🏗️ Architecture

PDF Upload → Azure Function → AI Processing (Azure OpenAI) → Cosmos DB Storage

✨ Features

PDF Text Extraction: Uses PyMuPDF to extract text from PDF resumes
AI-Powered Data Extraction: Leverages Azure OpenAI GPT-4o to extract:
- Personal information (name, email, location)
- Technical skills with proficiency levels and experience years
- Soft skills
- Work experience and current role
- Industry experience
- Certifications
Flexible Tagging: Support for custom tags (external, senior, remote, etc.)
Searchable Text Generation: Automatically creates optimized search text
Cosmos DB Integration: Stores structured data for efficient querying

📋 Prerequisites

Azure Subscription
Azure Functions Core Tools 4.x
Python 3.9+
Azure Cosmos DB account
Azure OpenAI resource with GPT-4o deployment

🚀 Quick Start

1. Clone and Setup

git clone <repository-url>
cd ResumeIngestor/Resume-Ingestor

2. Install Dependencies

pip install -r requirements.txt

3. Configure Environment Variables

Copy local.settings.json.template to local.settings.json and update the values:

{
  "IsEncrypted": false,
  "Values": {
    "AzureWebJobsStorage": "UseDevelopmentStorage=true",
    "FUNCTIONS_WORKER_RUNTIME": "python",
    "COSMOS_ENDPOINT": "https://your-cosmos-account.documents.azure.com:443/",
    "COSMOS_KEY": "your-cosmos-primary-key",
    "COSMOS_DATABASE_NAME": "exploredb",
    "COSMOS_CONTAINER_NAME": "resumes",
    "AZURE_OPENAI_ENDPOINT": "https://your-openai-resource.cognitiveservices.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2025-01-01-preview",
    "AZURE_OPENAI_KEY": "your-azure-openai-key",
    "AZURE_OPENAI_DEPLOYMENT_NAME": "gpt-4o",
    "AZURE_OPENAI_API_VERSION": "2024-12-01-preview"
  }
}

🔧 Environment Variables Configuration

Required Variables

Variable	Description	Example
`COSMOS_ENDPOINT`	Cosmos DB account endpoint	`https://mydb.documents.azure.com:443/`
`COSMOS_KEY`	Cosmos DB primary key	`ABC123...`
`COSMOS_DATABASE_NAME`	Database name	`exploredb`
`COSMOS_CONTAINER_NAME`	Container name	`resumes`
`AZURE_OPENAI_ENDPOINT`	Azure OpenAI endpoint URL	`https://myai.cognitiveservices.azure.com/...`
`AZURE_OPENAI_KEY`	Azure OpenAI API key	`XYZ789...`
`AZURE_OPENAI_DEPLOYMENT_NAME`	GPT deployment name	`gpt-4o`
`AZURE_OPENAI_API_VERSION`	API version	`2024-12-01-preview`

Getting Environment Values

Cosmos DB

Go to Azure Portal → Cosmos DB account
Keys section:
- COSMOS_ENDPOINT: Copy "URI"
- COSMOS_KEY: Copy "Primary Key"

Azure OpenAI

Go to Azure Portal → Azure OpenAI resource
Keys and Endpoint section:
- AZURE_OPENAI_KEY: Copy "Key 1"
- AZURE_OPENAI_ENDPOINT: Copy "Endpoint"
Model deployments section:
- AZURE_OPENAI_DEPLOYMENT_NAME: Your GPT-4o deployment name

🗄️ Cosmos DB Setup

1. Create Database and Container

# Using Azure CLI
az cosmosdb sql database create --account-name <cosmos-account> --resource-group <rg> --name exploredb
az cosmosdb sql container create --account-name <cosmos-account> --resource-group <rg> --database-name exploredb --name resumes --partition-key-path "/partition_key"

2. Container Configuration

Partition Key: /partition_key
Throughput: 400 RU/s (minimum)
Indexing Policy: Default (all paths indexed for text search)

📡 API Usage

Endpoint

POST https://your-function-app.azurewebsites.net/api/ingestresume

Request Format

{
  "FileUrl": "https://sharepoint.com/path/to/resume.pdf",
  "FileContent": "base64-encoded-pdf-content",
  "Tags": "external,senior,fullstack,remote"
}

Response Format

{
  "status": "success",
  "file_url": "https://sharepoint.com/path/to/resume.pdf",
  "tags": "external,senior,fullstack,remote",
  "extracted_text_length": 3000,
  "cosmos_document_id": "abc-123-def-456",
  "candidate_info": {
    "name": "John Doe",
    "email": "[email protected]",
    "location": "San Francisco, CA",
    "total_experience_years": 5,
    "current_role": "Senior Software Engineer",
    "technical_skills_count": 8,
    "soft_skills_count": 5,
    "certifications_count": 2,
    "industries": ["Technology", "Fintech"]
  },
  "message": "Resume processed and uploaded to Cosmos DB successfully"
}

📄 Document Schema

Documents stored in Cosmos DB follow this structure:

{
  "id": "unique-uuid",
  "partition_key": "active",
  "tags": "external,senior,fullstack",
  "personalInfo": {
    "name": "John Doe",
    "email": "[email protected]",
    "location": "San Francisco, CA"
  },
  "skills": {
    "technical_skills": [
      {
        "skill": "Python",
        "proficiency": "Expert",
        "years": 5
      }
    ],
    "soft_skills": ["Leadership", "Communication", "Problem Solving"]
  },
  "experience": {
    "total_years": 5,
    "current_role": "Senior Software Engineer",
    "industries": ["Technology", "Fintech"]
  },
  "certifications": ["AWS Certified Solutions Architect"],
  "searchable_text": "john doe python senior software engineer leadership technology",
  "metadata": {
    "fileUrl": "https://sharepoint.com/...",
    "filename": "john_doe_resume.pdf",
    "uploadTimestamp": "2025-01-18T10:30:00Z",
    "contentLength": 3000,
    "aiProcessed": true
  }
}

🔍 Querying Data

Basic Queries

-- Find all external candidates
SELECT * FROM c WHERE CONTAINS(c.tags, "external")

-- Search by skills
SELECT * FROM c WHERE CONTAINS(c.searchable_text, "python react")

-- Find senior candidates with specific experience
SELECT * FROM c 
WHERE c.experience.total_years >= 5 
AND CONTAINS(c.searchable_text, "senior")

-- Search by industry
SELECT * FROM c 
WHERE ARRAY_CONTAINS(c.experience.industries, "Technology")

-- Find candidates with certifications
SELECT * FROM c WHERE ARRAY_LENGTH(c.certifications) > 0

Advanced Queries

-- Multi-criteria search
SELECT c.personalInfo.name, c.experience.current_role, c.skills.technical_skills
FROM c 
WHERE CONTAINS(c.searchable_text, "python javascript") 
AND c.experience.total_years >= 3
AND CONTAINS(c.tags, "remote")

-- Skill proficiency search
SELECT * FROM c 
JOIN skill IN c.skills.technical_skills
WHERE skill.skill = "Python" AND skill.proficiency = "Expert"

🚀 Deployment

Local Development

# Start local development server
func start

# Test endpoint
curl -X POST http://localhost:7071/api/ingestresume \
  -H "Content-Type: application/json" \
  -d @test-payload.json

Deploy to Azure

# Login to Azure
az login

# Deploy function
func azure functionapp publish <your-function-app-name>

# Set environment variables in Azure
az functionapp config appsettings set \
  --name <your-function-app-name> \
  --resource-group <your-rg> \
  --settings \
  COSMOS_ENDPOINT="https://your-cosmos.documents.azure.com:443/" \
  COSMOS_KEY="your-key" \
  # ... other variables

🧪 Testing

Sample Test Payload

Create test-payload.json:

{
  "FileUrl": "https://example.com/resume.pdf",
  "FileContent": "JVBERi0xLjQKJdPr6eEKMSAwIG9iago8PAovVHlwZSAvQ2F0YWxvZwovT3V0bGluZXMgMiAwIFIKL1BhZ2VzIDMgMCBSCj4+CmVuZG9iago...",
  "Tags": "test,external,senior"
}

PowerShell Test

$body = @{
    FileUrl = "https://example.com/resume.pdf"
    FileContent = "base64-encoded-content"
    Tags = "test,external"
} | ConvertTo-Json

Invoke-RestMethod -Uri "http://localhost:7071/api/ingestresume" -Method POST -Body $body -ContentType "application/json"

🔧 Troubleshooting

Common Issues

1. "Cosmos DB connection settings not found"

Check environment variables are set correctly
Verify Cosmos DB account is accessible
Check firewall settings

2. "Azure OpenAI connection settings not found"

Verify Azure OpenAI resource is deployed
Check API key and endpoint
Ensure GPT-4o model is deployed

3. "Failed to parse AI response as JSON"

Check Azure OpenAI quota and limits
Verify deployment name matches
Review function logs for AI response details

4. PDF processing errors

Ensure FileContent is valid base64
Check PDF file is not corrupted
Verify PDF is not password protected

📞 Support

For issues and questions:

Create an issue in this repository

Name		Name	Last commit message	Last commit date
Latest commit History 20 Commits
.github/workflows		.github/workflows
.funcignore		.funcignore
.gitattributes		.gitattributes
.gitignore		.gitignore
README.md		README.md
function_app.py		function_app.py
host.json		host.json
local.settings.json		local.settings.json
requirements.txt		requirements.txt

Folders and files

Latest commit

History

Repository files navigation