Hi, my name is
Matteo Kitic.
I work on difficult problems and think about life.
I’m a graduate student enrolled at
About Me
Hello! I am Matteo, a full-stack developer and a dedicated Master's in Computer Science student enrolled at
I am passionate about leveraging technology to drive innovation and improve processes across various domains.
In my free time, I enjoy exploring new ideas and technologies, which fuels my commitment to personal and professional growth. I am excited about opportunities that allow me to contribute my skills and collaborate with others to create impactful solutions.
Where I’ve Worked
Analyst @ Warner Bros. Discovery
October 2022 -
- Collaborated with prominent brands such as MLB, NHL, NBA, & CNN to provide technical solutions for digital advertising products, contributing to eight figure revenue growth initiatives
- Developed dashboards for various brands' mobile applications to facilitate easy data visualization and analysis.
- Analyzed and delivered viewer insights to teams at prominent brands, optimizing the promotion, monetization, and overall execution of digital advertising products.
Some Things I’ve Built
Featured Project
Trading Bot
A Trading bot utilzing various different technical indicators and an ensemble learner to evaluate and execute trades.
- Python
- Matplot
- SciPy
- NumPy
- Pandas
Featured Project
Structural Analysis IoT
An IoT system utilizng a Raspberry PI and various sensors for visualizing real-time data and assessing strucural integrity using canny edge detection.
- Flask
- Python
- Pandas
- NumPy
- Matplot
- Dash
- OpenCV
- Raspberry Pi
Featured Project
Category Parition Method
Created a black box testing application for a client which uses equivalence class testing to minimize the number of possible test cases for an optimal runtime while maintaining reasonable test coverage. The application tackled this task by using the category partition method, implementing the shunting yard algorithm, and a proprietary expression evaluator.
- Java
- Spring
- Selenium
- Heroku
Other Noteworthy Projects
view the archiveMeasuring Bias in Machine Learning Models
Utilized statistical methods to determine the presence of a bias in a text classifier designed to detect toxicity in comments, categorizing them as 'toxic' or 'non-toxic'. This involved utilizing data analysis and statistical modeling to quantify potential biases. Specifically, calculating correlation coefficients among subgroups of protected classes against the text classifier's toxicity scores to assess any detected bias.
Ravens Matrices
Desgined an AI agent that's able to solve 3x3 and 2x2 Raven's Progressive Matrices. This was accomplished by employing a few different image processing techniques. It primarily works by using grayscale image manipulation along with metric-based methods and heuristics based methods.
Technical Indicators
Developed an application that foucsed on technical analysis and more specifically the implementation of different trading indicators and how they work.
Indicators analyzed:
- SMA
- Bollinger Bands
- Stocahstic Oscillator
- CCI
- Golden/Death Cross
Market Simulator
Developed a market simulator to execute trade orders, monitor dynamic portfolio valuation, and accounted for transaction costs to enhance financial accuracy.
Assessing Classification and Regression Trees
The research encountered in this project focuses on four superverised learning algorithms: decision tree learners, random tree learners, and two ensemble learners. The goal was to be able predict what the return for MSCI emerging markets index is by implementing these classification and regression tree (CARTs) algorithms
Portfolio Optimization
Portfolio optimization with Sharpe Ratio maximization. The primary objective is to determine the optimal allocation of funds to each stock in a portfolio. This optimization will be based on maximizing the Sharpe Ratio, which balances risk and return which allows investors to make informed decisions about their investments.
What’s Next?
Get In Touch
I am currently