Ph.D. candidate at Worcester Polytechnic Institute, Vernam Lab, advised by Prof. Berk Sunar.
On the industry job market — seeking research/engineering roles in trustworthy AI and hardware security.
- Trustworthy AI & LLM safety — guard-model robustness, jailbreaks, adversarial suffixes
- Hardware security — Rowhammer fault attacks, transient-execution vulnerabilities
- Post-quantum cryptography & FPGA-based hardware accelerators
- FAULT+PROBE — AsiaCCS 2025 (presented)
- Non-Halting Queries — SaTML
- LeapFrog — EuroS&P
Full list with TL;DRs and highlights: kemalderya.github.io/publications
- CVE-2024-5288 — wolfSSL: Rowhammer-based bit-recovery attack on fault-protected ECDSA in TLS 1.3 (disclosed as part of FAULT+PROBE)