Available for opportunities

Massimo Massetti

Offensive Security & AI Systems

I turn offensive research, AppSec triage, and AI systems into local-first security tools.

View Projects Get in touch
massimo@portfolio:~$
$ whoami
massimo — offensive security researcher
$ cat focus.log
HYDRA -> archived LLM honeypot research run
PDX -> Burp traffic into structured security deltas
BELIEF -> local reportability workbench
$ cat mindset.txt
validate impact, preserve evidence, automate noise away
$ █

01 About

Computer Science student at Université de Caen Normandie, focused on offensive cybersecurity and AI systems. Currently working through OffSec's PWK labs in preparation for the OSCP / OSEP path.

I actively hunt vulnerabilities across real-world platforms. Invited to private bug bounty programs on HackerOne and YesWeHack for major French groups in insurance, energy, finance, healthcare, and defense. Reported XSS vulnerabilities at Groupe Les Mousquetaires and session manipulation flaws at fintech companies. Selected writeups are redacted and focus on methodology, impact validation, and remediation rather than target-specific exploitation details.

Simultaneously working as an RLHF expert at Outlier, fine-tuning Large Language Models for alignment and robustness. I build the tooling around the work too: BELIEF for reportability triage, PDX for structured security deltas, and OSCP / OSEP tooling for disciplined local execution.

OSCP/OSEP In preparation
H1 / YWH Bug Bounty Platforms
3 Languages spoken
BELIEF AppSec Triage Engine

02 Projects

Updated Research Track
Open demo

HYDRA / PDX — Live Deception Layer

Controlled portfolio deception + security data pipeline

HYDRA began as an LLM-driven SSH research honeypot. The public VPS run is now archived, while this portfolio keeps a controlled deception layer for safe public demonstration. PDX remains the active local-first pipeline that normalizes scanner, decoy, and Burp signals into structured security events.

PDXactive focus
16Ddelta vector
Burpextension
7generators

The public surface is intentionally constrained: old static bait paths stay blocked, controlled decoys stay isolated, and the documentation explains the architecture without exposing routes, flags, or detection rules.

Read Documentation Burp Bridge GitHub
PythonJava MontoyaBurp Suite.pdxDataRouterMITREUnslothLoRA
B Public AppSec Tooling

BELIEF

Local reportability workbench for AppSec triage

Offline tool for turning scanner output, PDX deltas, and static-analysis findings into evidence-backed audit cases. The focus is conservative triage: what looks reportable, what is protected by guards, what is probably noise, and what still needs manual validation.

Localoffline triage
PDXsignal intake
SARIFinterop export
Reviewhuman decision

BELIEF is not an exploit generator or auto-submitter. It keeps the decision loop local and human-reviewed, with exports for JSON, SARIF, Markdown, and dataset-ready examples when the evidence is worth keeping.

GitHub
PythonAppSecBug BountyPDXSemgrepCodeQLSARIFZ3
⚔️ Latest Project

OSCP / OSEP Toolkit

Unified pentest workspace for OSCP / OSEP prep

After months of OSCP / OSEP preparation I noticed the bottleneck wasn't only the boxes — it was tool management. Built a single PyQt5 interface that syncs scope, credentials and notes across 80+ pre-configured tools. One LHOST setting propagates everywhere, commands auto-fill from a live scope vault, and embedded + external terminals coexist in the same window with the exam timer in the corner. 100% local — no cloud, no telemetry.

80+tools
2term modes
100%local
GitHub Read article
PythonPyQt5Kali LinuxWSLgOSCPOSEP
🏴‍☠️

Bug Bounty Hunting

Active hunter on HackerOne & YesWeHack with access to private programs across insurance, energy, finance, healthcare, and defense sectors. Identified stored XSS at Groupe Les Mousquetaires and session manipulation flaws at fintech companies. Methodology: recon, automation, manual exploitation, impact validation.

XSSSQLiSessionIDORCWE
LAB In Progress

Security Lab Infrastructure

Custom local lab for exploit development and defensive validation: memory corruption basics, controlled credential-protection experiments, Windows internals, detection notes, and malware-analysis-safe evasion research.

PythonCPyCryptodomeMinGW

03 Skills

> offensive

Pentesting Active Directory Vulnerability Research Bug Bounty Reportability Triage OWASP Top 10 Network Security

> development

Python C C# Rust PHP SQL Solidity Static Analysis

> tooling

Kali Linux Burp Suite Metasploit Wireshark Nmap Semgrep CodeQL SARIF Docker Git

> ai_research

RLHF LLM Security Model Alignment Fine-tuning Adversarial Testing PDX Deltas

04 Contact

Looking for an offensive security researcher, need a vulnerability assessment, or want to collaborate on AI security? Let's talk.

GitHub LinkedIn