Amine H

cybersecurity & Artificial Intelligence

CTM Business Solution • Temps plein

Mar 2026 - Present • 6 mos

Worked on securing industrial control systems (OT/ICS) within critical infrastructure environments, applying IEC 62443 standards — while integrating AI-powered tools to automate security workflows. • Conducted full security audits of industrial systems (OT/ICS), identifying critical vulnerabilities across network architecture, access control, and communication protocols. • Built an AI-assisted reporting pipeline using LLMs to automatically generate structured audit reports from raw technical findings — reducing report writing time by ~70% and ensuring consistent IEC 62443 compliance formatting. • Developed a Python automation script that parses security scan outputs and feeds them into an LLM prompt chain, producing actionable remediation recommendations tailored to each client's infrastructure. • Completed the AIR Cyber Bronze certification (industrial cybersecurity), gaining hands-on knowledge of OT threat modeling and risk assessment frameworks. • Designed reusable report templates (PDF/DOCX) auto-populated by AI from audit data, enabling the team to deliver client-ready documents within hours instead of days. Takeaway: bridging AI automation and industrial security — making complex technical audits faster, more consistent, and accessible to non-technical stakeholders.

Artificel Intelligence - Open source

Schneider Electric • Temps plein

Dec 2024 - May 2025 • 5 mos

Worked within Schneider Electric's AI Hub under Sylvain Marie, Senior Group Expert in Analytics & AI, contributing to production-level AI systems used at industrial scale. • Designed and integrated an AI pipeline component handling automated feature transformation, reducing manual preprocessing time by ~60% on internal datasets. • Built a reusable Python module for probabilistic data analysis, now integrated into the team's internal ML toolkit — improving model accuracy by 12% on anomaly detection tasks. • Implemented an API-based interface allowing non-technical teams to query ML models directly, removing the need for manual intervention and cutting response time from hours to seconds. • Collaborated with senior AI engineers on code review, model validation, and deployment readiness — gaining direct exposure to enterprise-grade AI workflows. Takeaway: learned how to ship AI solutions that are reliable, scalable, and usable by real teams — not just prototypes.