Gabi-comm/Network-Log-RAG-Chatbot

# Network-Log-RAG-Chatbot An intelligent log analysis system that uses embeddings to index network logs into a vector database (ChromaDB), allowing a local Language Model chatbot (Ollama) to query, analyze, and detect security anomalies. # Embedded Network Log Model with LLM Chatbot An intelligent, privacy-first network log analysis system. This project implements a Retrieval-Augmented Generation (RAG) pipeline that embeds raw network security logs, stores them in a vector database, and pairs them with a local Large Language Model (LLM) chatbot for interactive, natural-language security analysis. ## Features * **Log Embedding & Ingestion:** Uses Hugging Face embedding models to transform unstructured network logs into dense vector representations. * **Vector Storage (ChromaDB):** Stores embedded log data locally for rapid, semantic similarity search. * **Local LLM Integration (Ollama):** Queries the embedded logs using an open-source, local language model to ensure data privacy and security. * **Interactive Chat Interface:** Allows security analysts to query complex network behavior, trace anomalies, and generate incident summaries using natural language. ## Tech Stack * **Language:** Python 3.14+ * **Framework:** LangChain * **Embeddings:** Hugging Face (`langchain-huggingface`) * **Vector DB:** ChromaDB * **Local LLM Engine:** Ollama ## Project Structure The project is organized into modular scripts representing distinct phases of a Retrieval-Augmented Generation (RAG) pipeline: * **`HR_Policies.txt`**: The raw text dataset containing simulated policy information and hidden PII. * **`redact.py`**: Pre-processing script used to scan the raw text, detect sensitive information (like SSNs or emails), and clean it. * **`embed.py`**: Loads the sanitized data, splits it into digestible text chunks, generates vector embeddings, and stores them in ChromaDB. * **`retriever.py`**: Handles semantic search logic-taking a user query and fetching the most relevant document chunks from the vector database. * **`chatbot.py`**: The final execution script that wraps the retriever and links it to Ollama, providing an interactive natural language chat interface.