Bug Prediction
Throughout my pursuit of a Master's degree, I collaborated with Dr. Kontogiannis, who served as my supervisor at UWO, on an intricate bug prediction initiative. My thesis seamlessly merged the realms of Data Science, Software Engineering, and Machine Learning. In the course of this project, I meticulously engineered a database that interlinked code commits within extensive software projects with their corresponding JIRA issues. Subsequently, I adeptly extracted information and salient features from each commit, employing advanced analytical techniques to discern patterns indicative of potential bugs in the development pipeline. Upon completion, I successfully devised a commit transitions system model that yielded highly favorable results. Encouraged by my supervisor's suggestion, I am contemplating the prospect of formally submitting this accomplishment. This endeavor demanded a comprehensive grasp of the aforementioned disciplines, and I additionally delved into statistical analysis methodologies to ensure the attainment of optimal results.
Tools: Python, Pandas, Java, SQL, Excel, Matplotlib and Seaborn, SQLite, Jira, Jupyter Notebook, LateX, NumPy
Skills: Algorithms, Data Science, Database Management, Statistical Analysis, Project Management, Classification models and Clustering, Software Engineering, Feature Engineering, Data Visualization
Anomaly Detection Model
Within the context of an ongoing collaboration with a dedicated team at Caltech, my primary focus has been on the development of a sophisticated model designed to detect and subsequently predict anomalies within provided medical samples. The central objective of our partnership has been the creation of a data-generation model that leverages given inputs and the analysis of sampled data. This innovative approach is essential as the conventional method of scanning each sample individually would prove exceedingly time-consuming, rendering the experiment infeasible in practice. Considering that Artificial Intelligence is not within my primary area of expertise, the successful development of such a model and the achievement of the precise results desired by the medical team were notably impressive.
Tools: Python, Pandas, Excel, Matplotlib and Seaborn, Scikit-learn, Jupyter Notebook, Trello
Skills: Data Science, Image Processing, Model Development, Medical Domain Data Manipulation, Data Generation, Project Management, Data Visualization, Latex
Defect Prediction
For my Bachelor's thesis, I undertook a project with similarities to my Master's research. Employing the dataset provided by the PROMISE challenge in 2021, I developed an innovative classification model that demonstrated superior performance compared to established classifiers. Furthermore, during this endeavor, I conducted an array of rigorous statistical tests to thoroughly scrutinize the results. This experience greatly bolstered my confidence in the realm of statistical analysis. Notably, I was the sole undergraduate student within Dr. Miriyan's lab at Sharif University of Technology during this project. My efforts were recognized when my Bachelor's thesis was selected as the best of that term.
Tools: Python, Scikit-learn, Matplotlib and Seaborn, LateX, Git, PowerPoint
Skills: Data Analysis, Classification Models, Model Evaluation, Statistical Analysis, OOP, Data Visualization
Data Engineering
In a project for the Department of History at York University, I successfully created a comprehensive database from historical records of ships dating back to the 17th and 18th centuries. My responsibilities included developing the database structure, designing and implementing models and functions for efficient data retrieval, and creating a user-friendly information retrieval system for users without a background in computer science. The project involved extensive data cleaning to ensure accuracy and consistency, as well as identifying and merging identical entries. Additionally, I leveraged large language models to translate all the data from French to English, making the information accessible to a broader audience. This project not only required technical expertise but also a deep understanding of historical data and the ability to bridge the gap between complex data systems and end-users.
Tools: Python, Shell Script, PostgreSQL, DBeaver, Pyvis
Skills: Database Management, Information Retrieval, LLM, Data Engineering
Digital Design & Execution
Here, I’d like to share a bit about one of my passions that, while less directly related to Computer Science, has been an important creative outlet for me. Even before I began my secondary studies, I developed a strong interest in design and started using AutoCAD to bring my ideas to life. I applied AutoCAD across a wide variety of fields, including interior design, industrial design, jewelry design, and furniture design. One project was a custom-designed ring featuring an onyx stone set in silver; a piece where I was deeply involved in both the aesthetic and technical aspects of the design. I also created detailed industrial maps for different machine parts and even crafted intricate miniature 3D decorative models, blending precision with creativity. My passion for design even found its way into practical, everyday life; I designed a full set of cabinets using AutoCAD and then personally built them, using woodworking tools to turn digital plans into functional, real-world pieces. Through these experiences, I’ve not only honed my skills in technical drawing and design but also discovered how much I enjoy working at the intersection of creativity, engineering, and technology; an intersection that often complements my work in Computer Science.
Tools: AutoCAD, hand saws, power tools, sanding tools, 3D printing tools.
Skills: 3D Modeling and CAD Design, Industrial Design, Project Planning and Execution
BI Development
My time at ParsPooyesh Co. began as an intern, and I quickly became the youngest member of the team. Shortly thereafter, I was presented with an offer to join the company permanently—a proposition I regretfully declined to pursue my undergraduate studies, as I was just over 19 years old at the time. During my stint with the company, I made significant contributions in the realm of Data Engineering, predominantly utilizing GoLang. One of the standout projects, which harnessed the power of Go channels, involved developing a converter capable of transitioning between SQL and NoSQL data structures. This project ultimately proved to be a resounding success and was even commercialized by the company. Additionally, I was involved in a high-impact fraud detection project for the Telecommunication Company of Iran, demonstrating my proficiency in addressing complex challenges. Furthermore, I delved into the realm of Machine Learning to create an intelligent and personalized internet package recommendation system. My accomplishments in these projects, which aligned closely with the field of Business Intelligence, led to an offer to lead the Business Intelligence group within the company—a testament to the success and impact of my work in a relatively short span of time.
Tools: GoLang, SQL, NoSQL, MongoDB, TensorFlow, PowerBI, Tableau
Skills: Data Engineering, Business Intelligence, Concurrency and Parallelism, Fraud Detection, Applied ML, Project Management
Algorithmic Trading
Under the guidance of Dr. Habibi at SUT, I had the opportunity to craft a pioneering trading algorithm tailored specifically for Iran's Stock Market. What made this achievement noteworthy was its proven effectiveness, even during periods of market downturn when conventional trading algorithms were less reliable. It's worth noting that the Iranian economic landscape is characterized by a multitude of factors influencing the stock market, rendering standard international trading algorithms inadequate. Therefore, devising an algorithm that yielded favorable results during such circumstances was regarded as a significant success at the time. My proficiency in Data Analytics allowed me to navigate the complexities of the stock market and make a contribution to the field, despite my limited prior knowledge of stock trading.
Tools: Python, Django Framework, Git, GitLab
Skills: Algorithmic Trading, Financial Analysis, Project Management
Data Mining
I dedicated nearly a year to the lab of another distinguished faculty member at SUT, Dr. Fazli. During this time, I immersed myself in a complex Data Mining project of significant magnitude. The primary objective was to construct a comprehensive website that would provide intricate details regarding every transaction, price change, and the daily status of Iran's Stock market. A crucial component of this undertaking was the extraction of data from various online sources, necessitating an adept approach to web crawling. Leveraging the Django framework, I meticulously developed a series of web crawling techniques capable of traversing websites with varying degrees of data security measures. This experience not only enhanced my technical skills but also contributed significantly to the refinement of my abilities in collaborative teamwork.
Tools: Python, Django Framework, Git, GitLab, NodeJS, React
Skills: Data Mining, Web Crawling, Project Management
