Mhd Saria Al Lahham

AI Technical Lead, Cisco Systems · Montreal, QC, Canada
Ph.D., Queen's University

[ About | Research Interests | Education | Experience | Publications | Software ]

About

Senior AI/ML engineer with years of experience in Deep Learning, Generative AI, Large Language Models (LLMs), and Agentic Systems. I currently serve as AI Technical Lead (Operations and Development) at Cisco Systems in Montreal, where I work on Agentic AI for large-scale software systems: architecting enterprise-grade solutions that integrate LLMs with networking and telemetry data, and building automated root cause analysis tooling and multi-agent architectures for engineering teams.

Before Cisco, I was the AI R&D Manager at Apertera (formerly Alexa Translations), leading the design and deployment of agentic LLM systems, RAG pipelines, and MLOps workflows for professional machine translations, having joined as an AI Engineer. Earlier, I was an AI/ML Research Engineer at the Samsung Research America AI Center in Montreal, working on embedded AI, wireless communications (5G/6G), indoor localization, and human-state estimation.

I completed my Ph.D. in Computer Science at Queen's University in 2026; my thesis was on quantifying computational reliability in extreme edge computing, with applications to distributed edge AI. I also did my M.Sc. in Computer Science at Queen's University in 2022, on energy-efficient task allocation and incentive schemes for multi-orchestrator mobile edge learning, and my B.Sc. in Computer Engineering at Qatar University. Earlier still, I was a Research Assistant in the Smart IoT Systems Lab at Qatar University, designing AI-based protocols for ultra-reliable low-latency communication in smart health systems.

For a comprehensive list of my publications, please visit my Google Scholar profile.

Research Interests

• Large Language Models: Development and evaluation of multi-agent systems.
• Federated learning and distributed algorithms for the Internet of Things.
• Computational reliability for edge computing and AI.
• Reinforcement learning for wireless communication systems and network management.
• Wireless sensing: indoor human-state estimation, localization, and home automation.

Education

• Ph.D. in Computer Science, Queen's University, Kingston, 2026.
  Thesis: Quantifying Computational Reliability in Extreme Edge Computing: Applications to Distributed Edge AI.
• M.Sc. in Computer Science, Queen's University, Kingston, 2022.
  Thesis: Multi-Orchestrator Mobile Edge Learning: Designing Energy-Efficient Task Allocation and Incentive Schemes.
• B.Sc. in Computer Engineering, Qatar University, Doha, 2020.
  Graduated with High Order of Excellence.

Experience

• AI Technical Lead (Operations and Development), Cisco Systems, Montreal (Jul 2026 – present).
  Enterprise-grade agentic AI: LLMs integrated with networking and telemetry data, automated root cause analysis tooling, and multi-agent architectures deployed with reliability, observability, and low-latency guarantees.
• AI R&D Manager, Apertera (formerly Alexa Translations), Montreal (Jul 2025 – Jun 2026).
  System design and orchestration for production LLM and agentic workloads, and sovereign AI capabilities covering PII, data residency, and legal constraints on training, fine-tuning, and inference.
• AI Engineer, Apertera (formerly Alexa Translations), Montreal (Jul 2024 – Jun 2025).
  A RAG-powered translation system with agentic LLM orchestration that outperformed commercial translation engines on in-house data, with automated metric-collection and quality-evaluation pipelines.
• AI/ML Research Engineer, Samsung Research America, AI Center, Montreal (Oct 2022 – Feb 2024).
  Embedded AI on hardware and in simulation for indoor localization, computer vision, and RAN optimization; scientific papers and patent applications.
• Research Assistant, Smart IoT Systems Lab, Qatar University, Doha (May 2020 – Sep 2021).
  AI-based protocols for ultra-reliable low-latency communication (URLLC) in smart health systems.

Selected Publications

1. Reliable Federated Learning with Auction-Based Incentives at the Extreme Edge.
   IEEE Globecom 2024. Best Paper Award, IoT and Sensor Networks Symposium.
   [IEEE Xplore]
2. Multi-Agent Reinforcement Learning for Network Selection and Resource Allocation in Heterogeneous Multi-RAT Networks.
   [IEEE Xplore]
3. O₂AS: Optimistic Online Aerial Search via Vision-Language Models.
   [IEEE Xplore]
4. Probabilistic Mobility Load Balancing for Multi-band 5G and Beyond Networks.
   [arXiv]
5. Zero Touch Realization of Pervasive Artificial Intelligence as a Service in 6G Networks.
   [IEEE Xplore]
6. Reliable Federated Learning for Age Sensitive Mobile Edge Computing Systems.
   [IEEE Xplore]
7. Incentive-based Resource Allocation for Mobile Edge Learning.
   [IEEE Xplore]

A complete and up-to-date list is available on Google Scholar.

Software

• MERLIN: a library for simulating 5G wireless networks, with the ability to integrate and develop state-of-the-art reinforcement learning algorithms.
• 3GPP-chat: chat with 3GPP standards and documents locally, using Retrieval Augmented Generation (RAG).