Trustworthy investigative AI
Reliability, validation and accountable use of AI in evidence-sensitive workflows.
My work combines operational practice, technical development and applied research. The common thread is the reliability of decisions made around digital evidence: what can be preserved, what can be lost, what a tool changes, and how a conclusion can be documented and defended.
I work across digital investigations, digital forensics, cybersecurity and incident-oriented technical analysis. I am particularly interested in live systems, forensic acquisition, evidence preservation, traceability and the operational consequences of intervention.
The goal is not to present a catalogue of tools. It is to make technical choices understandable in terms of evidence, risk, footprint and reversibility.
I am completing a Master's degree in Computer Engineering, Cybersecurity and Artificial Intelligence at the University of Cagliari. My research interests include trustworthy AI for digital investigations, robustness of AI-assisted analytical workflows, provenance, privacy and technical accountability.
Reliability, validation and accountable use of AI in evidence-sensitive workflows.
Understanding where automated systems become unreliable and how those limitations should be documented.
Where automated assistance should stop and qualified human review must remain decisive.
Evidence First is an independently developed series of digital-evidence field guides built around state-dependent operational decisions rather than around products or command lists.
Evidence First — Vol. 1: Computer · Computer First Response & Forensic Acquisition · 2027 Edition.
SCOPE → OBSERVE → OPTIONS → GAIN → LOSS → FOOTPRINT → RISK → REVERSIBILITY → DECIDE → DOCUMENT.
The maintained publication site, current-edition information and companion materials live at evidencefirst.guide.
Only projects that support the professional and research narrative are emphasized here. Smaller experiments remain available through GitHub without being presented as equal-weight portfolio pieces.
Reproducible assessment of public-sector open data, API documentation, repository readiness and digital-governance risk.
Geospatial and event-oriented monitoring for public-source signals and situational awareness.
External exposure triage across DNS, TLS, email-security posture and public-facing technical signals.
This site is intentionally a curated professional portfolio. Public repositories and professional-network profiles provide external context for technical work and project history.