I developed a fraud-detection agent stack with LangChain and Langfuse, then tuned recursion depth, token use, and sampling settings to balance accuracy, cost, and latency under a tight contest budget.
Open to view full write-up and README details
LangChainLangfuseReActmulti-agentfraud detectiontoken budgetprompt tuning
I designed and benchmarked a multi-model defense pipeline combining FFNNs, GNNs, and transformers across malware triage, intrusion alerting, and tactic attribution. It highlights model behavior under noisy data and practical deployment trade-offs for SOC workflows.
Open to view full write-up and README details
MLPyTorchBERTGNNanomaly detectionmalware detectionintrusion detection
I investigated verifier edge cases and mapped bypass candidates to reproducible exploitability conditions in controlled labs. The work connects kernel internals with hardening guidance and realistic attacker paths.
Open to view full write-up and README details
kerneleBPFexploitLinuxverifier internalsXDPkernel hardening
I built an investigation workflow where AI-generated output must pass evidence-grounded checks before acceptance. The project emphasizes hallucination resistance, adversarial prompt robustness, and reproducible validation logic for forensic tasks.
Open to view full write-up and README details
AIforensicsadversarialgamehallucination defenseevidence validation
I processed 230k SSH honeypot sessions with NLP and sequence models to classify attacker intent and map behavior to ATT&CK techniques. The pipeline is built for analyst readability and downstream threat-intel use.
Open to view full write-up and README details
MLNLPhoneypotMITRE ATT&CKthreat intelligencesequence modeling
I implemented core cryptographic primitives and protocol exercises in C and Python, then stress-tested them through cryptanalysis and CTF-style break/fix challenges. It bridges formal concepts with hands-on attacker thinking.
Open to view full write-up and README details
CPythonOpenSSLCTFcryptanalysisapplied crypto
I engineered a secure timeout and watchdog mechanism on NXP S32K3 with FreeRTOS, validating behavior on hardware and in QEMU loops. The work prioritizes reliability under fault scenarios and secure embedded timing.
Open to view full write-up and README details
embeddedCNXPhardware securityFreeRTOSwatchdogQEMU
I designed and compared IPSec and TLS VPN lab topologies in SoftEther, then inspected traffic with Wireshark to validate handshake and tunnel behavior. The final analysis documents security and performance trade-offs.
Open to view full write-up and README details
VPNnetworkingWiresharkIPSecTLSnetwork forensics
I developed a security-first full-stack forum with RBAC authorization, JWT-backed sessions, and TOTP-based 2FA. Backend patterns were aligned to OWASP guidance with a focus on practical abuse resistance.
Open to view full write-up and README details
ReactNode.jsTOTPOWASPRBACJWTPostgreSQL
I co-built Budget HQ, a fintech-focused web platform created during the IEEE-HKN Global Hackathon where we secured 1st place worldwide. The delivery balanced rapid product iteration with secure full-stack engineering.
Open to view full write-up and README details
CTFhackathon1st placefull-stackfintechDocker