Principal AI engineer and researcher working on AI security, LLMs, and a bit of robotics. PhD in Busan, South Korea.
These days I lead AI research at SmartM2M (since Aug 2025), where I'm building SmartX, an AI penetration-testing platform driven by an automated LLM agent (SFT, RAG, and RL, running on-prem). I also work on LLM guardrails and automotive/cyber-physical security, and I lead a team of best AI engineers.
Before that, during my PhD at Pusan National University, I built CIPHER, a pentesting LLM that outperforms much larger models on hard HackTheBox machines, along with two frameworks (FARR and ASPIRE) for structuring security knowledge and evaluating agents.
A few other things I've spent time on:
- Reverse engineering. I showed that a DJI Mini drone could be hijacked mid-flight over Wi-Fi using a passive man-in-the-middle attack. Paper, demo, and code.
- Cryptographic engineering. I built HSM module protocols for OpenSSL, mbedTLS, WolfSSL, and GnuTLS.
- Side projects with a purpose: Shaikh, an realtime voice AI assistant for muroja'ah and grounded in Islamic references, and Kajian Maps, a live map of Islamic study sessions across Indonesia.
| Year | What |
|---|---|
| 2026 | Leading SmartX, an Anthropic's Mythos like on-prem AI pentesting platform |
| 2025 | Offensive AI and LLM red teaming at SmartM2M |
| 2024 | CIPHER, a pentesting LLM that beats models 10x its size |
| 2024 | Mid-flight DJI Wi-Fi hijack via passive MitM |
| 2023 | MECHA crypto API, 82.8% faster than standard context-switching |
| 2021 | COMET, tagged memory for RISC-V at under 1% overhead |
| 2014 | Won the Trinity Fire-Fighting Robot Contest |
AI and LLMs: Gemma, GPT-OSS, LLaMA, Mistral, Qwen, Gemini, Claude. For training and serving I use TRL, Axolotl, vLLM, and TGI, with LangGraph and SGlang for agents. Comfortable with fine-tuning (SFT, DPO), quantization, and RAG pipelines, including multimodal work with Whisper, ElevenLabs, and the usual diffusion stack.
Security: penetration testing, red teaming, LLM guardrails, CAN/ECU fuzzing, and cryptography (OpenSSL internals, ARIA/LEA/RSA). Familiar with OWASP Gen AI and MITRE ATLAS.
Systems: Python, C/C++, and Assembly are my main languages, with ROS, STM32, and RISC-V on the hardware side, and Docker, Nginx, and PostgreSQL for infrastructure.
- CIPHER: LLM for Guided Penetration Testing, Sensors 2024
- Behind the Wings: DJI Wi-Fi Hijacking, IEEE SecDev 2024
- COMET, Sensors 2021
- RIMI, IEEE TrustCom 2020
- MECHA, KCC 2023
The full list is on my Google Scholar.
- PhD, Information Convergence Engineering, Pusan National University (2023–2025)
- MSc, Information Convergence Engineering, Pusan National University (2020–2023)
- BEng, Computer Engineering, Sepuluh Nopember Institute of Technology (2011–2015)
If you're working on AI security, LLM evaluation, or robotics and want to compare notes, reach me at derryprata@gmail.com.