Inspiration
Since I'm in cyber security field .I was facinated with email analysis. I used to do header analysis which contained lot of information regarding an email. I learned lot of things regarding an email Analysis.
What it does
Analyzes email headers, validates SPF/DKIM/DMARC, extracts IOCs, detects phishing indicators, and generates AI assisted SOC investigation reports. which helps security analyst to verify with further investigation.
How we built it
I Developed using Python, FastAPI, Ollama, and Gemma 3 (4B), with ChatGPT as a programming assistant for development and feature implementation.
Challenges we ran into
Integrating the local LLM, handling varied email header formats, improving AI response accuracy, and ensuring reliable IOC extraction and authentication analysis. And programming challenges.
Accomplishments that we're proud of
Successfully integrated a local LLM with email security analysis to automate phishing detection and generate AI assisted SOC investigation reports.This project will help security analyst. In understanding the email analysis reducing their time.
What we learned
Learned how to integrate local LLMs, improve AI prompt design, parse email headers, and build scalable AI powered cybersecurity applications.
What's next for Kaaval
For future im planning to improve the investigation report. Also integrate it with threat intelligence.
Built With
- chatgpt
- fastapi
- ollama
- python
Log in or sign up for Devpost to join the conversation.