Advanced AI Form-Filling Automation Agent

This repository contains a Python-based automation agent designed to interact with online forms (e.g., Google Forms) and provide intelligent responses using a GPT-based model from Hugging Face. The agent mimics human behavior during form completion, making it ideal for advanced automation tasks that require thoughtful, AI-generated inputs.

Features

• Human-Like Form Filling: Automates form-filling tasks with randomized delays and human-like typing behavior.
• Hugging Face GPT-Neo Integration: Uses GPT-Neo (EleutherAI/gpt-neo-2.7B) to generate intelligent responses to form questions.
• Customizable Configuration: Easy to configure target form URLs and other parameters.
• Logging with Loguru: Provides robust logging to track the agent's actions and debug issues.

---

Components

1. Configuration

The program uses the following key configuration variables:

• TASK_URL: The URL of the form to automate (e.g., a Google Form).
• GECKO_PATH: Path to the GeckoDriver executable for Selenium.
• MODEL_NAME: Hugging Face model name (e.g., EleutherAI/gpt-neo-2.7B).

2. Dependencies

• Core Libraries: time, random, re
• Selenium: For browser automation.
• Loguru: For advanced logging.
• Transformers: For GPT-Neo model integration.

To install dependencies, use:

pip install selenium loguru transformers

3. AI Integration

• Model Initialization: The agent initializes a GPT-Neo model from Hugging Face:

  tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
  model = AutoModelForCausalLM.from_pretrained(MODEL_NAME)

• Text Generation: The call_huggingface_model function generates AI-driven responses to questions provided by the form:

  input_ids = tokenizer.encode(question, return_tensors="pt")
  output = model.generate(input_ids, max_length=100, temperature=0.7, num_return_sequences=1)
  answer = tokenizer.decode(output[0], skip_special_tokens=True)

4. Selenium Automation

• Browser Setup: The program uses Selenium to automate Firefox with GeckoDriver:

  service = Service(GECKO_PATH)
  driver = webdriver.Firefox(service=service)

• Form Interaction: Randomized delays and human-like typing simulate natural user interaction:

  def send_keys_human_like(element, text):
      for char in text:
          element.send_keys(char)
          time.sleep(random.uniform(0.02, 0.08))

5. Logging

Logs are generated using Loguru, making it easy to track execution and debug issues:

logger.add("data_input.log", rotation="10 MB")

Key logs include:

• Model loading and response generation.
• Browser actions and interactions.
• Error handling and troubleshooting.

6. Main Program Flow

The __main__ function provides an interactive CLI for the user:

• Start: Launch the automation process.
• Stop: Exit the program.

while True:
    cmd = input("\nType 'start' to run the automation, or 'stop' to exit: ").strip().lower()
    if cmd == "start":
        run_automation()
    elif cmd == "stop":
        print("Exiting the program.")
        break
    else:
        print("Unrecognized input. Type 'start' or 'stop'.")

---

Installation and Setup

1. Clone the repository:

git clone https://github.com/yourusername/ai-form-filling-agent.git

2. Install dependencies:

pip install selenium loguru transformers

3. Download GeckoDriver and set the GECKO_PATH in the script:

   • Download GeckoDriver
   • Update the path in the configuration section.

4. Update the TASK_URL to your target form.

5. Run the program:

python main.py

---

Troubleshooting

1. Hugging Face Model Freezing:

   • Use a smaller model (e.g., EleutherAI/gpt-neo-125M) for testing.
   • Ensure sufficient memory (16GB+ RAM recommended).

2. Selenium Issues:

   • Verify GeckoDriver is installed and GECKO_PATH is correct.
   • Ensure Firefox is updated.

3. Debugging Logs:

   • Check data_input.log for detailed execution logs.

---

Future Improvements

• Add support for other browsers (e.g., Chrome).
• Enable dynamic form parsing to handle complex forms.
• Add GPU support for faster Hugging Face model inference.