
CyberSec AI Research Assistant is an AI-powered threat intelligence tool built for cybersecurity professionals, SOC analysts, and security researchers.
π― The Problem
Security professionals waste valuable time manually searching across multiple sources for threat information during incidents. Switching between NVD, MITRE ATT&CK, threat feeds, and search engines is slow and error-prone β especially under pressure.
π‘ The Solution
One conversational interface that brings everything together instantly.
β¨ Key Features
- π Real-time cybersecurity threat research via web search
- π‘οΈ CVE vulnerability lookup from NIST National Vulnerability Database
- βοΈ MITRE ATT&CK technique lookup by ID (e.g. T1566, T1078)
- π§ Long-term memory using ChromaDB vector database
- π€ Model selector (GPT-4o-mini / GPT-4o)
- π Token usage and cost tracking per query
- π Security guard to block malicious queries
βοΈ How It Works
Users type a cybersecurity question into the chat interface. The LangGraph agent decides which tools to use β web search, CVE lookup, or MITRE ATT&CK β retrieves relevant information, and generates a comprehensive response. Past research is stored in ChromaDB and automatically retrieved for follow-up questions.
π οΈ Built With
LangGraph β’ LangChain β’ OpenAI β’ ChromaDB β’ Tavily β’ Streamlit β’ Python β’ NVD (NIST) API β’ MITRE ATT&CK API