Experience
AI Engineering Intern
Aramco — Aramco Research Center
Jun 2024 - Aug 2024
- Enhanced the synthetic data for a CNN industrial equipment classification model by creating 3D renders in Blender, improving lighting, customizing textures, and creating Python utility functions to streamline the process.
- Captured and processed 360-degree HDRi images, integrating them with the 3D renders, significantly boosting the model accuracy from 52% to 72%.
- Contributed to a project involving change detection of multi-spectral images by gathering datasets, testing models in PyTorch, and creating a flexible repository for testing different models.
- Created comprehensive documentation to ensure the smooth usage of my tools after my departure.
Undergraduate Course Assistant
UT Austin — Software Engineering Class
Jan 2024 - May 2024
- Supervised 6 student groups in developing full-stack websites, guiding them with project architecture design.
- Reviewed and provided feedback on +30 weekly student blogs, monitoring their progress and challenges, and reporting summaries to the professor.
- Conducted weekly office hours to assist students with understanding course concepts and fixing technical issues.
Research Intern
USC Viterbi School of Engineering - Data Science Lab
Jun 2022 - Aug 2022
- Implemented the Canonical Polyadic (CP) Tensor Decomposition algorithm using the Tensor Algebra Compiler library.
- Ran the CP algorithm on arbitrary data, compressing +90% of the data while maintaining its statistical significance.
- Presented the CP algorithm and the experiment results in a department-wide symposium.
Projects
The Word Engineer
Participated in a 3-month competition (SDAIA) to enhance ALLaM, a large Arabic language model, focusing on Arabic poetry generation improvements.
Key Features:
- Developed a semi-hardcoded Poem Analyzer to evaluate structure and linguistic accuracy.
- Designed a trial-and-error mechanism for iterative poem generation and error detection.
- Applied chain-of-thought techniques to improve coherence and plan verses beforehand.
- Built a Flask API for the model and a React front-end to display attempt history and feedback.
California Wildfires
Collaborated with a team of 5 to develop a React web application with a MySQL database, showcasing data about wildfires, nearby fire protection facilities, and California counties.
Key Features:
- Designed responsive React components using Figma and Amplify UI.
- Implemented efficient sorting, filtering, and searching functionality for +1400 elements.
- Built a multi-stage CI/CD pipeline to run and automate GUI acceptance tests with Selenium.
ML Energy Consumption Analysis Tool
Developed a Python-based tool using scikit-learn and CodeCarbon to analyze model accuracy vs. energy consumption during hyperparameter tuning.
Key Features:
- Implemented a script to systematically assess hyperparameter configurations across classifiers and datasets.
- Created automated visualization and analysis scripts, providing detailed performance and energy metrics.
- Presented findings in a paper:
Contact Me
Email me at
abdulrahman.alshahrani@kaust.edu.sa
or fill out the form below: