Imagine you are a Lead Developer or Data Analyst at a fast-growing startup. Every Monday morning, you face a data_dump<\/b> containing thousands of unstructured customer support tickets, tweets, and emails. Manually reading these to identify critical bugs versus minor feature requests is a bottleneck that delays product iterations. It is like trying to sort a mountain of mixed LEGO bricks by hand while more bricks are being poured on your head. This script acts as an automated sorting machine, reading every entry and filing it into the correct category with high precision.<\/p>\nThe solution<\/h2>\n
We leverage Python and the OpenAI API with Structured Outputs<\/b>. Instead of receiving messy text responses, we force the LLM to adhere to a specific Pydantic<\/b> schema. We use pathlib<\/b> for robust file system navigation to scan a directory of feedback files, process them through the model, and aggregate the results into a structured report.json<\/b>.<\/p>\nPrerequisites<\/h2>\n
Ensure you have Python 3.10 or higher installed. You will need an OpenAI API key. Install the necessary libraries using the following command:<\/p>\n
pip install openai pydantic python-dotenv<\/code><\/pre>\nThe code<\/h2>\n\n\"\"\"\n-----------------------------------------------------------------------\nAuthors: Sharanam & Vaishali Shah\nRecipe: Automated Feedback Classifier\nIntent: Use LLMs to convert unstructured text files into structured JSON data.\n-----------------------------------------------------------------------\n\"\"\"\n\nimport json\nfrom pathlib import Path\nfrom typing import List\nfrom pydantic import BaseModel\nfrom openai import OpenAI\nfrom dotenv import load_dotenv\n\n# Load environment variables from .env file\nload_dotenv()\n\n# Define the expected structure of our analysis\nclass FeedbackAnalysis(BaseModel):\n sentiment: str\n category: str\n priority: int\n summary: str\n\nclass FeedbackProcessor:\n def __init__(self):\n self.client = OpenAI()\n self.input_dir = Path(\"feedback_input\")\n self.output_file = Path(\"analysis_report.json\")\n \n # Create input directory if it does not exist\n self.input_dir.mkdir(exist_ok=True)\n\n def analyze_text(self, text: str) -> FeedbackAnalysis:\n \"\"\"Sends text to LLM and returns structured feedback.\"\"\"\n response = self.client.beta.chat.completions.parse(\n model=\"gpt-4o-mini\",\n messages=[\n {\"role\": \"system\", \"content\": \"Analyze the customer feedback and categorize it accurately.\"},\n {\"role\": \"user\", \"content\": text},\n ],\n response_format=FeedbackAnalysis,\n )\n return response.choices[0].message.parsed\n\n def process_all_files(self):\n \"\"\"Iterates through text files and aggregates results.\"\"\"\n results = []\n \n # Use pathlib to find all .txt files\n files = list(self.input_dir.glob(\"*.txt\"))\n \n if not files:\n print(f\"No files found in {self.input_dir}. Please add .txt files to process.\")\n return\n\n for file_path in files:\n print(f\"Processing: {file_path.name}\")\n content = file_path.read_text(encoding=\"utf-8\")\n \n analysis = self.analyze_text(content)\n \n # Combine original metadata with LLM analysis\n results.append({\n \"file_name\": file_path.name,\n \"analysis\": analysis.model_dump()\n })\n\n # Save the aggregated data to a JSON file\n self.output_file.write_text(json.dumps(results, indent=4), encoding=\"utf-8\")\n print(f\"Success! Report generated at {self.output_file}\")\n\nif __name__ == \"__main__\":\n processor = FeedbackProcessor()\n processor.process_all_files()\n<\/code><\/pre>\nCode walkthrough<\/h2>\n
The script begins by importing Path<\/b> from pathlib<\/b>, which provides an object-oriented approach to handling file paths, making the code cross-platform compatible without manual slash adjustments. We define a FeedbackAnalysis<\/b> class using Pydantic<\/b>. This class acts as a contract; it tells the OpenAI API exactly which fields we need (sentiment, category, priority, and summary) and their data types.<\/p>\n
The FeedbackProcessor<\/b> class handles the orchestration. Inside analyze_text<\/b>, we use the beta.chat.completions.parse<\/b> method. This is the modern way to handle LLM outputs because it eliminates the need for complex regex or error-prone json.loads()<\/b> calls. The API returns a validated object that matches our FeedbackAnalysis<\/b> model.<\/p>\n
The process_all_files<\/b> method uses the glob<\/b> function to find every .txt<\/b> file in the feedback_input<\/b> directory. It reads the raw text, passes it to the AI, and collects the structured objects into a list. Finally, we use write_text<\/b> to save the entire dataset into analysis_report.json<\/b>, ready for a dashboard or a database import.<\/p>\nSample output<\/h2>\n
When you run the script, the terminal will display the progress for each file found in the feedback_input<\/b> folder:<\/p>\n\nProcessing: ticket_001.txt\nProcessing: ticket_002.txt\nProcessing: tweet_99.txt\nSuccess! Report generated at analysis_report.json\n<\/code><\/pre>\nThe resulting analysis_report.json<\/b> will look like this:<\/p>\n\n[\n {\n \"file_name\": \"ticket_001.txt\",\n \"analysis\": {\n \"sentiment\": \"Negative\",\n \"category\": \"Bug Report\",\n \"priority\": 1,\n \"summary\": \"User unable to reset password via the mobile app.\"\n }\n }\n]\n<\/code><\/pre>\nConclusion<\/h2>\n
Automating data categorization with Python and LLMs removes the human fatigue factor from high-volume operations. By using Structured Outputs<\/b>, you transform an unpredictable AI into a reliable data pipeline component. This script can be scheduled as a cron job or integrated into a CI\/CD pipeline to monitor user sentiment in real-time, ensuring that critical issues never slip through the cracks again.<\/p>\n
\n\n\ud83d\ude80 Don’t Just Learn AI & LLMs \u2014 Master It.<\/h3>\n
\n This tutorial was just the tip of the iceberg. To truly advance your career and build professional-grade systems<\/strong>, you need the full architectural blueprint.\n <\/p>\n