About Me

I am a postdoctoral researcher at Khalifa University under the supervision of Prof. Merouane Debbah, working at the intersection of large language models and telecommunications. My research asks two questions: can LLMs solve hard telecom problems, and can telecom domain knowledge make LLMs better?

Recent Work

  • TelecomGPT — The first telecom-specific LLM in the world. A three-stage pipeline (continual pre-training, instruction tuning, alignment) that adapts open-source LLMs for telecommunications, outperforming GPT-4 on telecom-specific tasks using 7–8B parameter models.
  • RF-GPT — We propose the concept of Radio-Frequency Language Models (RFLMs) and take the first step with RF-GPT, which re-purposes vision-language models to reason about RF signals via spectrograms. 99.6% technology recognition across six wireless standards; competitive with CNN/ViT baselines at a fraction of the training cost.

Research Directions

LLM4Telecom — The broader vision is wireless intelligence: LLM-powered AI-native networks that autonomously understand, monitor, and manage the wireless environment. This spans RF signal understanding, spectrum awareness, semantic communications, and intelligent network orchestration.

Telecom4LLM — Applying principles from wireless and telecommunications to advance AI itself. This includes LLM quantization, edge deployment of LLMs under resource constraints, and leveraging communication-theoretic ideas to build more efficient and robust models.

I am also interested in developing theory and methods for semantic and goal-oriented communications — rethinking what gets transmitted by moving from bit-level fidelity to meaning-level and task-level fidelity.

Why This Blog?

This blog is a place to share our research — presenting work like TelecomGPT and RF-GPT in accessible long-form, beyond what a paper abstract can convey. It is also a way to document the rapidly evolving intersection of LLMs and telecommunications as both fields advance. Most importantly, I hope it helps build a community of researchers and engineers who see the same opportunity at this crossroads. If you are working on LLM+telecom, I would love to connect.

Contact

Feel free to reach out via GitHub or Google Scholar.