Computer Science (AI) graduate and video editor
British national, based in Cairo, Egypt, and open to relocating
I build machine-learning systems that understand video, and I produce the videos people watch. First-Class Honours graduate from the University of Hertfordshire, AWS certified, and running YouTube and social media for TraderLion & Deepvue since 2024.
I studied Computer Science with a specialism in Artificial Intelligence at the University of Hertfordshire and graduated with First-Class Honours. My coursework covered machine learning, computer vision, natural language processing, data science and software engineering, and my final-year project was among the highest-graded in my cohort.
Alongside the degree I built a second craft. Since February 2024 I've managed the YouTube channel and social media for TraderLion & Deepvue, a US trading-education company and its stock-research platform, editing broadcast-ready video in Premiere Pro and After Effects and shaping the content strategy.
I'm now looking for a role where I can put my technical background to work, along with the habit of delivering independently inside a remote, international team.
A system that watches a full match and finds the moments that matter. It extracts frame features with a ResNet backbone reduced through PCA, then uses temporal sequence models and segmentation to spot multiple classes of event across all 90 minutes.
A convolutional neural network that recognises everyday objects and clothing in one model. I merged CIFAR-10 with Fashion-MNIST, resizing and colour-matching the clothing images so both share one input format, then trained 11 convolutional layers with batch normalisation, dropout and data augmentation.
87%
overall accuracy on 20,000 unseen test images, across 10 CIFAR-10 classes and 10 Fashion-MNIST classes
Classifies students' social-media statements as normal or as showing signs of anxiety, depression or stress. After cleaning the text into bag-of-words features, I compared logistic regression with Naive Bayes, balanced the classes with SMOTE and tuned decision thresholds with precision-recall curves.
92.6%
accuracy on 7,957 unseen statements with logistic regression, the stronger of the two models
Java, JavaFX
Architected a system for speakers, attendees and scheduling, built on SOLID principles so it stays clean, maintainable and easy to extend.
Agents learn by trial and error to cross a 10×10 grid from one corner to the goal in the other, using Q-learning with epsilon-greedy exploration that tapers off as they learn. I compared a single agent with four agents sharing the grid, to see how competition changes the way they learn.
1,000
training episodes each for one agent on its own and for four agents learning side by side
One agent
Four agents
I run the channel end to end: editing, scheduling, channel branding and the overall content strategy, and I use viewer data to keep improving engagement and the production workflow.
As a freelance editor I've cut videos for industry experts including Eric Worre, Chase McAllister, Scott McPherson and Luke Matthews.
Launch films I've made for Deepvue's product releases. Browse through them, and tap or click a film to watch it with sound.
A 16-second launch spot: a wall of thousands of tickers collapses into one theme-filtered list, told almost entirely through motion and type.
How Deepvue's MCP server connects market data, screening, charts and alerts to AI assistants, told through live prompts and product UI.
Bringing indicators over from other platforms: the scripts convert, then plot straight onto Deepvue charts.
Launch film for Deepvue's community: posting and importing other traders' screens, watchlists and AI prompts.
Connect a brokerage, review every trade and track performance by emotion, all inside Deepvue.
One screen was never enough: Deepvue windows pop out and spread across every monitor on the desk.
Curated collections of historical chart setups, like high tight flags, for traders to study and build their own.
What clients have said about working with me.
Omar was a pleasure to work with, delivering all requested improvements swiftly – not to mention communication was a breeze and he paid attention to everything that was said. Highly recommend for any short-form video editing work.
Working with Omar was an absolute pleasure! He knows what he's doing, is able to create a unique style for your videos and always delivers! I can only recommend working with him!
Feb 2024 – Present
TraderLion & Deepvue, remote (US)
Channel I manage: youtube.com/TraderLion
Sep 2025
TMG Group, Cairo
2025, Hatfield, UK
University of Hertfordshire. First-Class Honours with a cumulative GPA of 4.36 / 4.5.
2021, Cairo
Egypt British International School. A* in Computer Science, Physics, Chemistry and English; A in Mathematics.
Amazon Web Services, Foundational
Amazon Web Services, Foundational
Hiring for an engineering or AI role, or need a video editor? Email me or book a free 30-minute call.
omourad713@gmail.com