AquilaX Securitron

AquilaX's flagship AI Model

AquilaX Securitron Training Loop

AquilaX powers across two core domains:

1. All in one scanner: Utilizing eight security scanners for code base assessment.

2. AI Engine Securitron: A highly trained model specialized in application security (AppSec).

Securitron has been trained on over 300 million open-source projects, learning from their source code and identified security vulnerabilities. It has also been meticulously trained on the triaging work performed by leading cybersecurity engineers. As a result, Securitron is the first AI engine on the market extensively trained on billions of data points, enabling it to reason like a human security expert.

This dataset includes:

1. Source Code snippets

2. Identified vulnerabilities

3. Triage results from top security engineers

4. Tags categorizing findings as:

   • False Positive

   • False Negative

   • True Positive

   • True Negative

   • Undefined

This dataset is pivotal in training our Securitron model. Securitron plays a crucial role in assessing new vulnerabilities, distinguishing between False and True Positives, offering an unparalleled contextual understanding.

Moreover, Securitron features a Chat component serving as both API and UI. This allows AquilaX users to interact, gain insights into vulnerabilities, understand why certain findings are True Positives, and learn effective mitigation strategies.

Continuous training of Securitron ensures it evolves with each new triage performed by our security engineers, facilitating ongoing improvements in our system.