AI/ML Researcher & Engineer
I’m happiest when learning something completely new and turning it into a working system.
BIOGRAPHY & RESEARCH FOCUS
I received my Ph.D. in Electrical Engineering from the New Jersey Institute of Technology (2026) under the supervision of Prof. Joerg Kliewer. My research focused on reliable AI, machine learning, and information retrieval under uncertainty, with an emphasis on robust algorithms for decentralized learning, semantic retrieval, and distributed computation.
During my Ph.D., I completed a research internship at Futurewei Technologies, where I developed machine learning and optimization methods for energy-efficient resource allocation in next-generation communication networks.
Prior to that, I received my M.Sc. in Electrical and Electronics Engineering from Özyeğin University (2021) under the supervision of Prof. Murat Uysal. My master's research focused on wireless communication systems, visible light communications, and Simultaneous Wireless Information and Power Transfer.
I have published research in leading IEEE journals and conferences, including IEEE Transactions on Wireless Communications, IEEE Transactions on Information Forensics and Security, IEEE GLOBECOM, IEEE ISIT, and IEEE ITW. I also serve as a reviewer for IEEE Transactions on Communications, IEEE Transactions on Wireless Communications, IEEE Transactions on Mobile Computing, and IEEE Wireless Communications Letters. My current interests include reliable machine learning, systems large language models, retrieval-augmented generation (RAG), and AI agents.
LATEST UPDATES
Received my Ph.D. in Electrical Engineering from the New Jersey Institute of Technology.
Our work, “Context-Aware Search and Retrieval Under Token Erasure,” was presented at the 2026 Information Theory and Applications Workshop in San Diego, California.
Completed a research internship at Futurewei Technologies, where I worked on machine learning and optimization methods for energy-efficient resource allocation in next-generation communication networks.
THEORETICAL FOUNDATIONS & IMPACT
My research has evolved from communication systems and mathematical optimization to machine learning, reliable AI, information retrieval, and modern AI systems. Throughout this journey, I have been motivated by developing mathematically grounded methods for machine learning, information retrieval, and reliable AI.
My research journey began with communication systems and wireless networks, where I studied visible light communications, resource allocation, and optimization techniques for next-generation communication systems. This work provided a strong mathematical foundation that continues to shape how I approach machine learning and AI research today.
During my Ph.D., my research expanded toward reliable AI and machine learning, with an emphasis on decentralized learning, robust computation, and learning under uncertainty. I developed algorithms that improve the reliability and resilience of AI systems operating in challenging environments.
My recent work focuses on information retrieval and semantic communication, developing optimization-based methods for embedding-aware retrieval, semantic search, and efficient information transmission under communication constraints.
Building on my research background, I have recently been expanding into modern AI systems, including large language models, retrieval-augmented generation (RAG), AI agents, and AI engineering. I enjoy applying strong mathematical and machine learning foundations to rapidly learn emerging technologies and build practical AI applications.
PEER-REVIEWED PAPERS
S Ghasvarianjahromi, J Barr, Y Yakimenka, J Kliewer
arXiv preprint arXiv:2604.18424
S Ghasvarianjahromi, A Kiani, A Xiang, J Kaippallimalil, T Saboorian, ...
IEEE Networking Letters
S Ghasvarianjahromi, Y Yakimenka, J Kliewer
2025 IEEE Information Theory Workshop (ITW), 821-826
S Ghasvarianjahromi, Y Yakimenka, J Kliewer
IEEE Transactions on Information Forensics and Security
M Bakshi, S Ghasvarianjahromi, Y Yakimenka, A Beemer, O Kosut, ...
2024 IEEE International Symposium on Information Theory (ISIT), 2502-2507
S Ghasvarianjahromi, Y Yakimenka, J Kliewer
GLOBECOM 2023 - 2023 IEEE Global Communications Conference
M Uysal, S Ghasvarianjahromi, M Karbalayghareh, PD Diamantoulakis, ...
IEEE Transactions on Wireless Communications 20 (10), 6715-6728
S Ghasvarianjahromi, M Karbalayghareh, PD Diamantoulakis, ...
2019 IEEE 30th annual international symposium on personal, indoor and mobile
S Ghasvarian, J Abouei, M Azadi, A Anpalagan
2019 27th Iranian Conference on Electrical Engineering (ICEE), 1321-1326
SOFTWARE & DATASETS
• Fine-tuned a DistilBERT model for sentiment classification on the SST-2 dataset using Low-Rank Adaptation (LoRA) and the PEFT framework, reducing the number of trainable parameters to approximately 0.8% of the original model. • Built an end-to-end fine-tuning pipeline using the Hugging Face ecosystem (Transformers, PEFT, Datasets, and Evaluate), including data preprocessing, model training, evaluation, checkpointing, and reproducible experimentation. • Published the fine-tuned model and accompanying model card to the Hugging Face Hub, demonstrating reproducible model sharing and deployment.
• Built a local-first, multi-agent job search assistant that plans targeted searches, retrieves postings from multiple sources, ranks candidate-job fit, and generates tailored application materials. • Designed a modular pipeline using Google’s Agent Development Kit (ADK), with specialized agents for search planning, multi-source retrieval, match scoring, job-fit analysis, resume alignment, application drafting, and notifications. • Implemented configurable YAML-based profiles, weighted job-matching heuristics, URL-based deduplication, SQLite persistence, CLI controls, and optional Gemini-powered reasoning with deterministic offline fallbacks.
ACADEMIC DEGREES
New Jersey Institute of Technology, Newark, NJ
Dissertation: 'Computing and Retrieval under Uncertainty: Reliable Decentralized Computation and Importance-Aware Semantic Communication'. Advised by Prof. Joerg Kliewer.
Ozyegin University, Istanbul, Turkey
Thesis: 'Simultaneous Lightwave Information and Power Transfer in Underwater Visible Light Communications'. Advised by Prof. Murat Uysal.
Yazd University, Yazd, Iran
Thesis: 'Improvement of Routing Algorithms in Diffusion-based Molecular Communication'. Advised by Dr. Jamshid Abouei.
New Jersey Institute of Technology, Newark, NJ, USA
Fully funded Ph.D. research assistantship covering tuition and stipend in the Department of Electrical and Computer Engineering.
The Scientific and Technological Research Council of Turkey (TÜBİTAK)
Competitive graduate scholarship supporting M.Sc. studies and research in Electrical and Electronics Engineering.
Özyeğin University, Istanbul, Türkiye
Merit-based full scholarship awarded for graduate studies in Electrical and Electronics Engineering.