About Me

Overview

During this contract role, I worked as a Project Coordinator, overseeing projects from Initiation & Execution to Closure. I worked with a range of stakeholders to prioritise project progress to meet deadlines & requirements. As an Engineer, Project Management was a completely new world I entered — I spent the first 2–3 months adapting to this new type of role, completing project management trainings and familiarizing myself with the company's workflow.

While I cannot give details into the projects I had the opportunity to work on, I can share that I was involved in finance, engineering, software, business development and IT projects, each presenting with different challenges based on number of stakeholders involved, the involvement and commitment needed to ensure progress, company-set priority and so on.

I would describe this role as a fast-paced one, being proactive was non-negotiable in this environment. My days were spent with constant followup with multiple stakeholders, multi-tasking projects to stay aware of updates & changes, which would help tweak project plans, schedules and strategies, all to ensure a project is always moving on time or ahead of schedule within our realm of control.

I took away many lessons form this experience, aside from the importance of proper project planning, process adherence

Key Achievements

• Managed portfolio of 6 engineering and IT projects ($2M+ total value) using Scrum framework and Jira, facilitating sprint planning and tracking deliverables to ensure on-time completion.
• Led project planning and business case development for securing $1.5M in funding for a Maintenance Training Program from local government stakeholders.
• Recovered 6 weeks of schedule delays by resolving 35+ cross-stakeholder issues and managing dependencies across engineering, IT, and operations teams.
• Documented project outcomes and lessons learned from 3 completed project phases, identifying process improvements that reduced future planning cycles by ~10% of project length.

Overview

During my 8-month Co-Op with Miru, myself and another Co-Op were responsible for managing the Testing Laboratory and process flows from device receiving to failure analysis. Our team was the backbone of the company — we tested each new device we received, which was the output of research conducted by our scientists, to supply them with critical data regarding performance with changing formulations, parameters and materials. During my tenure, Miru was still in their "research phase", therefore my work in the laboratory was the backbone of the company at the time, highlighting the criticalness of my job to be performed effectively.

Key Achievements

• Established a systematic failure analysis methodology on 200+ electrochromic devices using optical microscopy and electrical testing; identified root causes (delamination, electrode degradation) for 85% of failures.
• Redesigned testing procedures by optimizing parameters (voltage profiles, test cycles), improving test efficiency by 30%, enabling R&D to test 10 additional device configurations daily.
• Prepared technical documentation for failure investigations; collaborated with R&D engineers to recommend design modifications based on failure trend analysis, reducing defect rates by 15%.
• Analyzed test data from 400+ devices to identify performance trends and failure patterns, reducing transcription errors by 90% and documentation time by 40%.

Overview

[Overview of the role — team, company context, what you were brought in to do.]

Key Achievements

• [Achievement #1]
• [Achievement #2]
• [Achievement #3]

Overview

[Overview of the role.]

Key Achievements

• [Achievement #1]
• [Achievement #2]

Overview

This new project involves learning more about textile engineering through MIT's online course, "Innovations in Textile Engineering: Fibers, Yarns, Nonwovens, & More" to get a better grasp on the textile sciences field, along with exploring & researching about performance textiles of interest such as GoreTex, Polartec and so on.

Will give a proper update soon!

Key Achievements

• [Achievement #1]
• [Achievement #2]
• [Achievement #3]

Overview

Developing a Wearable SOS Signal Transmitter Device for use in rural mountain sports. The idea arose after I fell into a small ditch while skiing at Mount Seymour in March 2025. I was lucky not to sustain serious injuries, but it forced me to wonder what would have happened had I been skiing alone. This sparked research into modern solutions — from avalanche beacons to RECCO tags — and revealed a gap between effective and widely accessible mountain safety tools.

I began researching sensor options: IMU, EKG, and pressure sensors as primary candidates, with barometric, EMG and temperature sensors as secondary considerations. I started with pressure sensors using Velostat — a material whose resistance changes directly with applied pressure — constructing samples and analysing incoming data on an Arduino board.

I created a few smple armbands with a few pressure sensors sewed into the armband and quickly fund out the __ (all over the palce) readings due to the sensitivity of the velostat and the construction of the sensor was not great... This made me reconsider the effectiveness and usefulness of pressure sensors for this project and its purpose. Pressure sensors are local body sensors here... Therefore I decided to not move ahead with pressure sensors for the project, they could be a project of their own as they are quite a complex research topic.

I decide to focus on the IMU sensor, it gave me acceleration, jerk, angular velocity and euler angle data - more than enough parameters to work with. After reaching this step, I decided to then focus on the second main function of the project, the SOS signal transmission. At first, I defined the concept of an SOS message, it must contain certain details about the user, primarily their personal detials, location for first responders to use and possibly any known medical conditions and even environmental conditions to aid with search & rescue. This meant I needed a few extra modules for my board, a GPS, a trasmitter & reciever system and a relevant environmental sensor.

The main focus was the communication system that needed to be used. There are many ways data can be communicated but only a few useful methods in a given rural environment, assuming no cellular connection would be available, which the options turn to be, some form of long range RF communication (excluding WiFi and Bluetooth) or a Satellite communcaition. Both paths were viable, where a satellite system would, at first glance, seem like the ideal solution for rural environments, however a quick Google search of such modules made me eliminate this option due to costs of $300+ and additional subscription fees with RockBlocks Satellite system. I did eliminate this option for prototype development, but kept it in the back of my mind for a possible future consideration for the final product.

That left long range RF communication, which I quickly came across LORA RF modules (Low-Power Radio Frequency), with preliminary search showing a range of 5-10km with typical modules, which was a promising final piece to this puzzle I am forming. Of course I would need to conduct further research into how these devices operate and conduct practical range tests. Range is highly dependent on a number of factors, primarily being physical interference between the sender and receiver modules, urban area would show worse performance, an ideal would be an open area, possible even better when sent from a high elevation. LoRa devices also need to have their paramters configured, such as trasmisison power, spreading factor coding rate and more. Changes in these factors are what will be tested against LoRa range performance and observations in signal quality.

Moving forward, I began protoypting using the following modules: ESP32 as the MC to program, a MPU6050 as my IMU, BME 280 for temperature and pressure sensing, EKG for heart rate monitoring, SD card writer and RYLR998 LoRa module. What I eventually came up with is shown in Figure 3, my first finalized prototype breadboard ready to be used to collect data while skiing, powered by a power bank. I came up with a testing emthodology and trialed it at home to test rectify any potential issues with my code and setup to ensure data integrity for a few hours.I took it to Lebanon where I recorded a total of a 2 hour ski session, classifying the following events: "Normal Skiing", "Chairlift" &, "Walking".

Key Achievements

• [Achievement #1]
• [Achievement #2]
• [Achievement #3]

Overview

As part of my final year capstone project at UBC, I had the great opportunity to be mentored from the Canadian Space Agency for building a Pressure Vessel based on the several set requirement from the CSA & based on the lunar conditions of the Moon.

Key Achievements

• [Achievement #1]
• [Achievement #2]
• [Achievement #3]

Overview

Every time we hear a familiar voice, our brain either recognizes the voice or registers it as a new one. This project asked whether our brain recognition system leaves a detectable trace in our brainwaves, and whether a machine learning model could learn to read it.

Working in a team of three for MANU 465, we conducted data collection on 30 participants and had each of them listen to a series of 30 second audio clips while wearing a Muse 2 EEG headband, which is a commercial device that reads electrical activity from the brain's surface. For each clip, participants indicated whether they recognized the voice or not, giving us 180 labeled recordings to work with.

From there, we processed the raw brainwave data and extracted hundreds of signal features, which are patterns in how different frequency bands behaved & responded, how signals varied over time, and how activity differed across sensor locations on the scalp. These features were fed into a range of machine learning models, from simple classifiers like Naive Bayes and K-Nearest Neighbors, to more complex approaches including neural networks and a CNN trained on visual representations of the brainwave signals.

The honest takeaway was that 180 samples is a tough starting point for this kind of problem. Most models learned the training data well but struggled to generalize, causing a major overfitting issue that was unavoidable with a small dataset. The best-performing approach hit 80% accuracy under cross-validation, which was encouraging, but the results also made it clear what the next steps to improving our research was having more participants, better hardware, and smarter feature selection informed by neuroscience rather than brute-force quantity.

The project my first introduction and application to machine learning and was a lesson in the realities of working with small, noisy biological datasets as it was in machine learning and that turned out to be just as valuable.

Key Achievements

• Designed and ran a full human-subjects data collection experiment using EEG hardware across 30 participants
• Built an end-to-end ML pipeline from raw brainwave signals to trained classifiers, evaluating 7+ models
• Identified key limitations and proposed a clear roadmap for improving generalizability in future iterations

Overview

Nuclear power is one of the cleanest and most energy-dense sources of power we have, and demand for it is only growing. But behind every reactor is an enormous amount of precision engineering, namely including thousands of welds on the pipes that keep the reactor core cooled. A single weld failure in that environment isn't just a maintenance issue. This project was about understanding how to get those welds right.

Working in a team of six for MTRL 472, we used Simufact Welding, an industrial FEA weld simulation tool, to model how different welding parameters affect the structural integrity of stainless-steel pipe joints. The core question was whether MIG welding, which is faster and cheaper, could be a viable alternative to TIG welding in nuclear piping applications, where TIG has traditionally been the standard.

We modelled two 316L stainless steel pipes being butt-welded together, testing 15 different simulation runs that varied welding speed, current, voltage, and clamping force across both MIG and TIG processes. From each run we extracted temperature distributions, residual stresses, pipe displacement, and weld fusion profiles.

The results pointed clearly to welding speed and current as the dominant variables. Slower speeds meant more heat input, wider HAZ, and higher residual stresses, all negative effects for long-term corrosion resistance. Our base MIG case (Run 3) came out as the most balanced: complete weld penetration, residual hoop stresses well within the safe threshold for 316L steel, and a faster process than TIG. The best TIG run produced slightly lower and more stable residual stresses overall, but at meaningfully higher cost and time, being a trade-off when a nuclear plant involves upwards of 30,000 individual welds.

The project also had its honest limitations. Single-pass welds only, a coarse mesh to keep simulation times manageable, and some software instability that constrained how far we could push the model. But as a study in what simulation-driven materials analysis actually looks like in an engineering context, it was a genuinely useful exercise to analytically understand the differences and limitations between different welding techniques.

Key Achievements

• Built and ran 15 welding simulations in Simufact across MIG and TIG processes, systematically varying speed, current, voltage, and clamp force
• Estimated labour cost implications across both processes, showing MIG welding could save several million dollars over the lifetime of a nuclear plant build
• Determined MIG Case 3 as the optimal paramter set, thus improving my udnerstanding the effects of different weld paramaters based on the 3 welding technqiues investiagted

Overview

As a Lebanese, moving to Canada for the first time was a big challenge to overcome, the cultural differences, lifestyle, and most importantly, being far away from what I call home. During ym first few years at university, I was happy to be able to befriend many people from a variety of cultures and backgrounds, but I always felt my Lebanese identity did not have a hook in Vancouver or at UBC. Thats when I decided to pursue creating an Arabs club in my final year at the UBC to help solve this issue for many students like myself.

After forming a team of a few known interested people, we set off making many events around promoting our cultural identity while forming a solid community on campus for students alike and any others interested in our diverse cultures.

Events ranged from food oriented events, music events, outdoor events, charitable workshops and support circles during difficult periods for our community.

Key Achievements

• Raised over $6,000 through various fundraisers to support charities by aiding communities in need in Lebanon & Palstine
• Organized numerous community and cultural events to promote awareness of our diverse cultures to create a sense of communit and belonging on the UBC campus
• Established a club with over 100 paid members...

Overview

[What this volunteer role involved.]

Key Achievements

• [Achievement #1]
• [Achievement #2]