AI-powered platform for detecting and analyzing toxic communication patterns. Built with FastAPI, PostgreSQL, pgvector, HuggingFace, OpenAI GPT-4o-mini, and RAG. Features PII anonymization, fine-tuned DistilBERT gate, zero-shot classification, and Responsible AI design.
python nlp text-classification postgresql openai cyberbullying-detection rag content-moderation fastapi responsible-ai streamlit transformers-models pii-anonymization coercive-control gaslighting llm generative-ai langchain pgvector toxic-communication
-
Updated
Jul 1, 2026 - Python