Embedded devices and IoT endpoints are becoming the weakest link in today’s supply‑chain‑driven software ecosystems. To protect these systems, security analysts must repeatedly reverse‑engineer the firmware, identify processor architectures and locate vulnerabilities.

In the first part of the talk I will give a clear, non‑technical introduction to firmware analysis: what firmware is, why its security matters, and the main steps of firmware analysis.
The core of the presentation demonstrates how data‑science and AI techniques can support and automate these steps. For example, how a small neural network is used to identify processor architecture, or how software‑component and version detection can be improved by data‑analysis techniques.
I will also outline the first results of AI agents usage in finding metadata for software components and creating self‑assessment reports.

By the end of the session attendees will see concrete workflows that combine classic reverse‑engineering with modern machine‑learning pipelines, demonstrating how AI‑driven automation accelerates security and compliance self‑assessment in embedded and industrial environments.