Mohammad Bazzal

-Led a team of private tutors delivering personalized instruction in AI, Machine Learning, NLP, and Computer Science fundamentals. -Supervised and mentored fellow tutors, ensuring high-quality, hands-on teaching and student success. -Helped multiple students graduate with strong AI foundations and real-world coding skills. -Delivered tailored lessons on machine learning algorithms, data analysis, and NLP applications. -Empowered students to code, think critically, and even publish research or projects.

→ Invented a decoder-free minimum distance estimation algorithm for turbo/convolutional codes, achieving 100× speedup over the state-of-the-art TIM baseline on K=512 bit frames, directly applicable to 5G/LTE and satellite link design (published IEEE VTC2025). → Designed DPCCC, a novel parallel concatenated code architecture that outperforms 5G reference configurations across all evaluated frame sizes; validated via large-scale C/MATLAB simulation pipelines (published ISTC2025). → Designed graph-based tree-search algorithms and automated full-family LTE turbo code benchmarking using Python and C. → Proposed interleaver design improvements combining joint optimization and distance-centric search; submitted to IEEE Open Journal of the Communications Society (OJ-COMS), 2026.

→ Developed multi-agent actor-critic deep RL algorithms (PyTorch) for beamforming optimization in next-generation MIMO wireless environments, demonstrating measurable gains over baseline heuristics. → Modeled non-Markovian radio environments and implemented centralized-training / decentralized-execution (CTDE) frameworks applicable to real-time 5G resource management. → Bridged academic RL research and production-grade simulation; contributed to the team's ongoing publication pipeline on intelligent wireless systems.