Final-year Digital Security engineering student at EURECOM. Reverse engineering, malware analysis, and applying machine learning to detection and triage.
Seeking a six-month cybersecurity internship outside France from September 2026. EURECOM provides the convention de stage.
meerkat — SOC alert triage
Normalizes Wazuh, Suricata and AMiner into one ranked daily queue with MITRE ATT&CK context. The queue is built to cover as much distinct attack activity as a fixed daily budget allows. Ten cases a day reach 58 of 60 attack steps on the AIT-ADS benchmark, against 19 for native detector severity. Technical report.
Python scikit-learn Wazuh Suricata AMiner MITRE ATT&CK
malfamily — malware classification
Classifies malware into six behavioural families using assembly instruction mnemonics as the sole feature source, across PE, ELF and Mach-O. 77.9% accuracy against a 25.3% baseline, 0.77 macro-F1 over 472 samples.
Python Ghidra scikit-learn x86
mnemocrypt-enhanced — cryptographic function detection
Improved an IDA Pro plugin that misread compression code as cryptography. Rebalanced its training data and added lane-aware SIMD instruction counting. False positives on held-out goodware 48 → 9, precision 0.60 → 0.86.
Python IDA Pro x86-64 scikit-learn
smudgeC — C source-to-source obfuscator
Written in pure C with a hand-built lexer and symbol table, no parser generator or external dependency.
C
ctf-writeups — 25+ walkthroughs
Binary exploitation, forensics, cryptography and web, across picoCTF, Root-Me and TryHackMe.
| Reverse engineering | Ghidra, IDA Pro, gdb, x86-64 / ARM / RISC-V assembly |
| Defensive security | Suricata, Wazuh, AMiner, MITRE ATT&CK, Wireshark, nmap |
| ML and data | scikit-learn, pandas, NumPy, matplotlib |
| Languages | C, Python, OCaml, Java, SQL |
jiacheng.wang@eurecom.fr · Paris, France · Mandarin, French, English (TOEIC 930)