Project Information
- Category: Machine Learning & Data Analysis
- Client: Zweig Group
- Date: 2024
- Technologies: Django, GPT-4o-mini, Python, OpenAI API
- Role: AI Engineer & Full-Stack Developer
SAL: Sentiment Analysis Lens
A Django app that analyzes employee feedback from "Best Firms to Work For" surveys using fine-tuned GPT-4o-mini models, providing sentiment classification, response summarization, and key theme extraction for participating companies.
Key Features:
- Sentiment Classification: Categorizes responses as Positive, Negative, or Outliers
- Category Summarization: Generates summaries for six survey categories using fine-tuned models
- Overlap Detection: Identifies responses related to multiple categories
- Strengths & Weaknesses Analysis: Extracts top three positive and negative themes
- User-Friendly Interface: Frontend displaying sentiment breakdowns and insights