About me

I am Nguyen Phuc Tran, a software engineer and applied AI/ML researcher who builds production-grade systems and turns network data into decisions. Over the past decade I have shipped software across the full stack, led teams, and now design and deploy AI/ML and Generative AI solutions for mobile communication networks as a Ph.D. candidate at Concordia University (Montreal, QC, Canada), working with industry partners Ericsson and Ciena.

I work at the intersection of three disciplines:

• Software Engineering — designing, building, and operating reliable systems end to end (backend, frontend, DevOps) with a strong emphasis on quality, testing, and maintainability.
• Data Science — modeling network behavior, predicting KPIs, and extracting insight from large-scale, real-world industrial datasets.
• Applied AI / ML / Generative AI for Telecom — bringing LLMs and machine learning into 5G/B5G network operations: resource allocation, energy efficiency, service assurance, and root cause analysis.

What I do

Applied AI / ML / Generative AI for telecom networks

I design end-to-end pipelines that put modern AI to work on real network problems. Recent work includes:

• LLM-augmented root cause analysis (RCA). Built and evaluated Fine-Tuning, RAG, and Hybrid LLM approaches to construct an RCA knowledge base from industrial support tickets — cutting analysis time by roughly 30% and improving network resilience. (IEEE Access, 2026)
• Energy-aware LLMs. Designed a pipeline that trades off energy efficiency against model performance for fault-ticket analysis, using quantization and pruning to reduce energy consumption while improving accuracy on real-world telecom datasets. (ICECET, 2025)
• ML-based KPI prediction for 5G/B5G. Developed machine learning models to estimate throughput and derive downstream KPIs for end-to-end network slices, supporting SLA-driven service assurance. (EuCNC/6G Summit, 2023)
• AI for service assurance & optimization. Implemented AI-driven algorithms that optimized network resources by up to 54.85% and reduced operating costs in 5G/B5G network-slicing simulations.

Data science

I turn large, messy, real-world network and operational data into models and insight — KPI prediction, throughput estimation, mining support tickets, and quantifying trade-offs (e.g., energy vs. performance). My work runs on production-scale industrial datasets, not toy benchmarks, and is validated against operational metrics.

Software engineering

Before (and alongside) research, I spent 5+ years as a senior / full-stack software engineer across telecom and healthcare platforms:

• Built and maintained backend services, web and mobile apps, and DevOps pipelines for a healthcare PaaS and clinic-management platform (Silentium / Egopulse).
• Developed high-availability cloud core-middleware for Ericsson on an OpenSAF-based architecture (DEK Technologies).
• Built real-time dashboards and profiling software for industrial process control (KIC).
• Practiced test-driven development, code review, CI/CD, and microservice architecture throughout.

Technical skills

Research focus

I am pursuing my Ph.D. and working as a graduate research assistant under the supervision of Prof. Brigitte Jaumard and Prof. Tristan Glatard in the Large-Scale Optimization Lab at Concordia University. My research applies artificial intelligence — including large language models — to mobile communication networks, with a focus on resource allocation, energy efficiency, green mobile networks, system design, root cause analysis, and system optimization.

Let’s connect

I am always open to collaboration on applied AI/ML for networks, software engineering, and data-driven systems.

• 💼 LinkedIn
• 💻 GitHub
• 📄 Google Scholar
• 📑 Publications · CV
• ✉️ Email