I am a Software Engineer with experience working in an intensive startup environment alongside an AGILE team. I have almost three years of experience in object-oriented programming, particularly in the development of high-volume, low-latency products. My professional track record highlights my proficiency in tackling intricate data-intensive applications, with a keen focus on optimizing efficiency.
Outside of my professional commitments, I enjoy going on long hikes in nature, reading books and working on research projects.
View Resume
GPA: 3.6/ 4.3
Notable Achievements: Received a scholarship for Fin-ML Research and was hired three times for distinct programming projects, two of which were used to facilitate course capstone projects.
Coursework: Decision Analysis, Machine Learning, Machine Learning II: Deep Learning, Adv. Statistics, Forecasting Methods, Algorithms, Complex Networks Analysis, Optimization, Statistical Modelling
GPA: 8.4/ 10
Notable Achievement: One of only two teams in my cohort to receive research funding.
Coursework: Data Structures, Object Oriented Programming, Data Management Systems, Distributed Systems, Operating Systems, Adv. Applied Mathematics I/ II/ III/ IV, Digital Logic Design, Microprocessor, Circuits, Web Technologies
at Shift-Technology, a Fintech Unicorn Startup
Tech Stack: C#, T-SQL, Python
Monitoring: Grafana
CI/ CD: TeamCity, Octopus
Note: This position did not encompass analytics, which is typically a core aspect of Data Science; instead, it focused solely on backend tasks.
Used design patterns in C# to make a generic and reusable data transformation tool for the Fraud Detection product, resulting in better scalability and enhancing efficiency for more than 10 clients.
Enhanced company-wide impact by improving data transformation efficiency by 5% through the utilization of better data structures and adjustments to mapping algorithms.
Coded hash-based deduplication filters and processes to help efficiently ingest and process tables with up to 9 billion highly duplicated rows of data within server memory constraints.
at MKSoftware, a company that makes ERP solutions
Main Product: DocuFire
Tech Stack: C#, React.js, Typescript, T-SQL
CI/ CD: Windows Installer, Installshield
Achieved a 15% increase in efficiency by implementing asynchronous parallel processing of pages in the pdf document generation module of the core ERP Solution allowing for processing of documents at scale.
Led the transformation of a legacy project's data access layer into a REST API, preserving all existing methods. Strengthened the system with improved unit tests, ensuring enhanced functionality.
Designed and developed the MVC backend for an interactive ASP.Net Core Web Application, seamlessly integrating it with the REST API mentioned above.
at HEC Montréal, University of Montréal; developing web application to
facilitate course
capstone projects
Tech Stack: C#, MySQL, Python
Created and hosted a scalable C#/ASP.Net Core website with MySQL for the 'Sports Analytics' course. The system fulfilled functional requirements such as user registration, team and game sign-ups, and included an on-server Machine Learning simulation.
Developed Python visualizations for data from a sports simulation ML model, subsequently adapting them for use on the server of the website.
at Sentometrics, writing a Machine Learning research paper
Tech Stack: R, Python
Used topic modeling (R) to analyze and gain insights about financial research over time.
Scraped data, preprocessed text, and annotated the text corpus.
Created and transformed several covariates for 33 Financial Journals.
Implemented, optimized, and compared multiple Structural Topic Models for relevant insights.
at ERPSim Lab, HEC Montréal
Tech Stack: Python, SAS
Documenting and verifying validity (Python) of simulated data for use in Cortex - a platform for learning and practicing data science concepts.
Programming in SAS and writing guides for students and teachers about using SAS-EM effectively to analyze the simulated datasets.
Created a complex road network of bike sharing service usage by overlaying it with a road network of the city.
Converted data into daily and weekly demand time series and incorporated conversion of time series data into complex networks using distance metrics and also a visibility graph.
Analyzed networks using global/local network descriptors, & community detection algorithms.
The project features a GUI for visualizing convex hull algorithms using Python & Tkinter.
Users can explore algorithms like Brute Force, Jarvis' March, Graham Scan, and Divide & Conquer by generating random points.
A specialized LeftHull algorithm is included for polygon convex hulls.
These algorithms are implemented from the psuedocode in their original research papers, wherever available.
We classify images of recyclable trash into different categories by using the following:
Support Vector Machines (SVM): One vs Many
Convolutional Neural Networks (CNN):
Architectures: VGG10, VGG13, VGG16, ResNet34
Hyperparameters: Pooling, Batch Normalization, Depth, Data Augmentation
in The Journal Of Economic Surveys
Employed an unsupervised Machine Learning Bayesian Model, the Structural Topic Model, to analyze financial journals.
in the 2nd International Conference on Trends in Electronics and
Informatics, published by IEEE
Developed an Arduino project integrating ultrasonic sound sensors and a gyroscope onto a glove, accompanied by custom code aimed at enhancing navigation for individuals with visual impairments.
in the 2017 International Conference on Intelligent Sustainable
Systems, published by IEEE
Created an Android application designed to issue natural disaster warnings and guide users to the closest warehouse, factoring in considerations such as distance, capacity, and other relevant metrics.
in the 2017 2nd International Conference on Communication and Electronics
Systems, published by IEEE
This project entails the development of a web application aimed at modernizing and optimizing record management processes in colleges and universities. By digitizing these tasks, it effectively lessens the administrative burden. The application encompasses features such as an online paper-checking module, attendance tracking, and a digital notice board, all of which contribute to enhanced efficiency in academic operations.