OpenAI API Configuration

The LinkedIn Job Analyzer uses OpenAI’s GPT models to analyze job descriptions and generate structured summaries. This guide covers API key setup, model selection, and cost management.

Getting Your OpenAI API Key

Create OpenAI Account

Visit OpenAI Platform and create an account if you don’t have one.

You’ll need to provide a phone number for verification.

Add Payment Method

Navigate to Settings → Billing and add a payment method.

OpenAI’s free trial has limited credits. You’ll need to add a payment method for production use.

Generate API Key

Go to API Keys in your dashboard
Click Create new secret key
Give it a descriptive name (e.g., “LinkedIn Job Analyzer”)
Copy the key immediately - it won’t be shown again

The key will start with sk-proj- or sk-

Set Usage Limits (Recommended)

In Settings → Billing → Usage limits, set:

Monthly budget cap (e.g., $10)
Email notifications at 75% and 90%

This prevents unexpected charges.

Configuring the API Key

Environment Variable Setup

Add your API key to the .env file in the project root:

.env

OPENAI_API_KEY=sk-proj-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Security Best Practices:

Never commit .env files to version control
Don’t share API keys in screenshots or logs
Rotate keys if accidentally exposed
Use separate keys for development and production

How the Application Uses the API Key

The application loads the API key at startup (gpt_analyzer.py:14-23):

from dotenv import load_dotenv
import os
from openai import OpenAI

load_dotenv()
api_key = os.getenv("OPENAI_API_KEY")

if api_key:
    client = OpenAI(api_key=api_key)
else:
    print("Warning: OPENAI_API_KEY not found in .env file")

If the API key is missing, the application will start but return a warning message when attempting to analyze jobs.

Model Selection

The application currently uses GPT-3.5 Turbo by default (gpt_analyzer.py:63):

response = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[...],
    temperature=0.7,
    max_tokens=500
)

GPT-3.5 Turbo vs GPT-4

Feature	GPT-3.5 Turbo	GPT-4
Speed	Fast (~2-3 seconds)	Slower (~10-20 seconds)
Cost per 1M tokens	$0.50 input /$ 1.50 output	$30 input /$ 60 output
Quality	Good for straightforward tasks	Better reasoning, more accurate
Best for	High-volume analysis	Complex requirements analysis

Changing the Model

To use GPT-4 or other models, edit inteligencia_artificial/gpt_analyzer.py:63:

# Options:
model="gpt-3.5-turbo"      # Default - fast and cheap
model="gpt-4o-mini"        # Smaller GPT-4, good balance
model="gpt-4"              # Most capable, expensive
model="gpt-4-turbo"        # Fast GPT-4 variant

The application limits input to the first 30 skills (gpt_analyzer.py:41) to control token usage and costs.

Cost Considerations

Token Usage Per Request

Each job analysis uses approximately:

Input tokens: ~300-500 (prompt + job description)
Output tokens: Up to 500 (configured limit)
Total: ~800-1000 tokens per analysis

Estimated Costs

Using GPT-3.5 Turbo:

Per job analysis: ~ $0.001 -$ 0.002 (0.1-0.2 cents)
100 job analyses: ~ $0.10 -$ 0.20
1,000 job analyses: ~ $1 -$ 2

Using GPT-4:

Per job analysis: ~ $0.03 -$ 0.06 (3-6 cents)
100 job analyses: ~ $3 -$ 6
1,000 job analyses: ~ $30 -$ 60

Cost Control Tips:

Start with GPT-3.5 Turbo for testing
Set monthly budget limits in OpenAI dashboard
The app limits skills to 30 items to reduce costs
Monitor usage in OpenAI dashboard regularly

API Usage Limits

Rate Limits

OpenAI enforces rate limits based on your account tier:

Tier	Requests per Minute	Tokens per Minute
Free Trial	3	40,000
Tier 1 ($5+ spent)	3,500	200,000
Tier 2 ($50+ spent)	3,500	450,000

For normal usage (a few analyses per minute), you won’t hit these limits. Batch processing might require delays between requests.

Handling Rate Limits

If you encounter rate limits, the application will return an error. Consider:

Add retry logic with exponential backoff
Implement delays between batch requests
Upgrade account tier by spending $5+ to increase limits

Error Handling

The application handles common API errors (gpt_analyzer.py:75-77):

Missing API Key

Warning: No se ha configurado la API Key de OpenAI. 
Crea un archivo .env con OPENAI_API_KEY=...

Solution: Add the API key to your .env file

API Request Errors

Error al consultar ChatGPT: [error message]

Common causes:

Invalid API key
Insufficient credits/quota exceeded
Network connectivity issues
Rate limit exceeded

Troubleshooting Steps

Verify API Key

Check that your key is correctly set:

python -c "import os; from dotenv import load_dotenv; load_dotenv(); print(os.getenv('OPENAI_API_KEY')[:20] if os.getenv('OPENAI_API_KEY') else 'Not found')"

Test API Key

Test the key directly:

curl https://api.openai.com/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Check OpenAI Dashboard

Visit the OpenAI Platform to:

Verify billing status
Check usage and limits
Review error logs

Check Application Logs

The application prints debug information:

[IA] Generando resumen con ChatGPT...

Look for error messages in the console output.

Security Best Practices

Protecting Your API Key:

Store keys in .env files, never in code
Add .env to .gitignore
Use separate keys for dev/staging/production
Rotate keys regularly
Enable usage alerts in OpenAI dashboard
Never log or display full API keys

Key Rotation

If your key is compromised:

Immediately revoke it in OpenAI dashboard
Generate a new key
Update your .env file
Restart the application

Environment-Specific Keys

For production deployments:

# .env.development
OPENAI_API_KEY=sk-proj-dev-key...

# .env.production
OPENAI_API_KEY=sk-proj-prod-key...

This allows tracking usage separately per environment.

Monitoring Usage

OpenAI Dashboard

Monitor your usage at platform.openai.com/usage:

Daily/monthly token consumption
Cost breakdown by model
Request counts and error rates

Application Logging

The analyzer logs each request (gpt_analyzer.py:38):

[IA] Generando resumen con ChatGPT...

You can enhance logging to track:

Number of tokens used per request
Response times
Error rates

Testing the Configuration

Verify your OpenAI setup is working:

# test_openai.py
from inteligencia_artificial.gpt_analyzer import AIAnalyzer

analyzer = AIAnalyzer()
result = analyzer.generar_resumen(
    titulo="Software Engineer",
    habilidades=["Python", "Flask", "REST APIs", "Docker"]
)

print(result)

Run the test:

python test_openai.py

You should see a structured analysis of the job requirements.

Next Steps

Basic Usage

Start analyzing LinkedIn job postings

Architecture Overview

Understand how the AI analyzer integrates

Get Started

Features

Configuration

Getting Your OpenAI API Key

Configuring the API Key

Environment Variable Setup

How the Application Uses the API Key

Model Selection

GPT-3.5 Turbo vs GPT-4

Changing the Model

Cost Considerations

Token Usage Per Request

Estimated Costs

API Usage Limits

Rate Limits

Handling Rate Limits

Error Handling

Missing API Key

API Request Errors

Troubleshooting Steps

Security Best Practices

Key Rotation

Environment-Specific Keys

Monitoring Usage

OpenAI Dashboard

Application Logging

Testing the Configuration

Next Steps

Basic Usage

Architecture Overview

Build docs developers (and LLMs) love

Get Started

Features

Configuration

​Getting Your OpenAI API Key

​Configuring the API Key

​Environment Variable Setup

​How the Application Uses the API Key

​Model Selection

​GPT-3.5 Turbo vs GPT-4

​Changing the Model

​Cost Considerations

​Token Usage Per Request

​Estimated Costs

​API Usage Limits

​Rate Limits

​Handling Rate Limits

​Error Handling

​Missing API Key

​API Request Errors

​Troubleshooting Steps

​Security Best Practices

​Key Rotation

​Environment-Specific Keys

​Monitoring Usage

​OpenAI Dashboard

​Application Logging

​Testing the Configuration

​Next Steps

Basic Usage

Architecture Overview

Build docs developers (and LLMs) love

Getting Your OpenAI API Key

Configuring the API Key

Environment Variable Setup

How the Application Uses the API Key

Model Selection

GPT-3.5 Turbo vs GPT-4

Changing the Model

Cost Considerations

Token Usage Per Request

Estimated Costs

API Usage Limits

Rate Limits

Handling Rate Limits

Error Handling

Missing API Key

API Request Errors

Troubleshooting Steps

Security Best Practices

Key Rotation

Environment-Specific Keys

Monitoring Usage

OpenAI Dashboard

Application Logging

Testing the Configuration

Next Steps