Alex Sciuto

Masters of Data Analytics Student | UCF

Focus in Machine Learning & Data Science

About Me

I am a data science and machine learning enthusiast with a passion for using technology to solve real-world problems. My goal is to use my skills in data analysis and model building to drive results in industries that can benefit from the insights generated from data. With this portfolio, I hope to showcase my abilities and demonstrate my commitment to becoming a professional data scientist or machine learning specialist. From automating property management tasks to creating predictive models for various industries, I am eager to tackle challenging projects and make a meaningful impact with my work.

Portfolio Projects

Gen3-PokeGAN

Is a project where I implemented CycleGAN architecture to transform modern Pokémon images into classic Generation 3 pixel art style, demonstrating my skills in machine learning and image processing. I developed and trained the models using PyTorch and created an interactive web app with Streamlit for real-time transformations. This project highlights my expertise in GANs, model deployment, and software development.

If live demo is asleep, click the “Yes, get this app back up!“ button

Bird Call Classifier

Is a project where I developed multiple machine learning model to classify bird species based on their calls. I processed raw audio data into structured representations, extracted features using the VGGish model, and trained classifiers like Random Forest, SVM, and Neural Networks. This project highlights my skills in audio data processing, feature extraction, and applying machine learning models for classification tasks.

github-RAGchain

Is an innovative tool I developed that combines LangChain, vector stores, OpenAI embeddings, and Retrieval-Augmented Generation (RAG) to analyze and understand GitHub repositories. I created a conversational AI interface using Streamlit and OpenAI's language models, allowing users to explore complex codebases through intuitive chat commands. This project showcases my expertise in AI, natural language processing, and software development for enhancing code comprehension and interaction.

If live demo is asleep, click the “Yes, get this app back up!“ button

Digital Democracy API

is a FastAPI application I developed to summarize legislative bills and generate reports with pros and cons based on the bill's content, leveraging the OpenAI API for summarization. This project highlights my skills in API development, web scraping, and natural language processing, enabling efficient extraction and summarization of legislative information. The API fetches bill details from the Florida Senate website, generates comprehensive summaries and pro/con analyses, and creates detailed PDF reports, demonstrating my expertise in integrating multiple technologies for impactful solutions.

Exploring Martin Dockery’s ‘Inescapable’: A Journey Through Time with OpenAI Embeddings

is a project where I utilized OpenAI’s embeddings model, text-embedding-ada-002, to analyze the script of Martin Dockery's play 'Inescapable' through vector space conversion and clustering. By transcribing audio recordings and segmenting the text, I performed a semantic analysis that visualized the repetitive nature and thematic elements of the play. This project demonstrates my ability to apply advanced natural language processing techniques and data visualization for insightful literary analysis.

Training nanoGPT to Channel Nietzsche: A Journey into AI-Driven Literature

is a project where I trained Andrej Karpathy’s nanoGPT model on Friedrich Nietzsche’s "The Will to Power" to generate text in his unique philosophical style. By preprocessing Nietzsche’s aphorisms, encoding them, and configuring the model for efficient training, I demonstrated my skills in natural language processing and fine-tuning GPT models. This project showcases my ability to leverage AI to emulate complex literary styles, highlighting both the potential and limitations of AI-driven literature.

Education & Relevant Experience

01. Education

  • Aug 2017 -Dec 2020 , Tampa, Florida, United States · The University of South Florida

  • August 2023 - Present,Orlando, Florida, United States | The University of Central Florida

    Relevant Coursework:

    ‣ Machine Learning

    ‣ Data Mining Methodologies 1 & 2

    ‣ Parallel & Distributed Databases

    ‣ Network Science

    ‣ Statistical Analysis

02. Work Experience

  • Jan 2024 - Present, Orlando, Florida, United States · Remote

    Developing an API that implements AI functionalities to Florida Senate Bills and federal legislation extracted from FLSenate.gov. Implementing the API into the Digital Democracy App so voters can have a better understanding of legislation passed when voting for representatives. The system is run within an AWS EC2 container using FastAPI, and it makes robust use of OpenAI API models. Skills and responsibilities include:

    ‣ Developing a RESTful API using FastAPI to process legislative data and integrated it into the Digital Democracy Website. Ensured seamless data flow and user experience.

    ‣ Applying machine learning models and OpenAI API to analyze legislative text. Enhanced the accuracy and depth of insights provided to users.

    ‣ Deploying and managing the API within an AWS EC2 container. Ensured the system's scalability and reliability.

  • Mar 2020 - Present · 4 yrs, Orlando, Florida, United States

    Management of leasing, marketing, price determination, and upkeep of over a dozen Real-Estate Investments Portfolios (i.e., 250+ rental properties) utilizing data driven approaches. Skills and responsibilities include:

    ‣ Utilizing statistical methods to determine rental listing prices from numerous variables data mined from the Stellar MLS (e.g., location, square footage, & other leased properties).

    ‣ Data wrangling & analyzing data in R and python from rental payment ledgers in order to automate the generation of documents (e.g., lease agreements & 3-day notices) with Google Scripts.

    ‣ Communicating with real-estate investors to make data-driven decisions to secure returns such as: selecting prospective tenants, finding cost-effective vendors, and determining when eviction filings are necessary

  • Mar 2022 - Aug 2022 · 6 mos, Orlando, Florida, United States

    Assessment of prospective clients ability to qualify for a residential mortgage, and assistance with helping them through the process from application approval to closing. Skills & Responsibilities include:

    ‣ Generating email campaigns in order to obtain prospects and referrals, in addition to employing data driven methods to improve efficacy such as A/B testing.

    ‣ Assessing prospect client application data (e.g., bank statements, monthly expenses, & employment) in order to create a strategy to purchase a home while maintaining financial security.

  • Aug 2017 - Dec 2019 · 2 yrs 5 mos, Tampa, Florida, United States

    Administration and scoring of psychological tests to diagnose developmental and learning disabilities (e.g., ADHD, ASD, ODD, Depression, ect.) for diagnosis for kids (ages 4-17). Skills and responsibilities include:

    ‣ Administration & scoring: the Kaufman Test of Educational Achievement (KTEA-3), Social Language Development Test (SLDT-A), Conners Kiddie Continuous Performance Test (K–CPT 2), and several more.

    ‣ Writing patient background reports following intake sessions that include family, birth, medical, psychiatric, educational, and cognitive history and confirming diagnoses from administered tests and the DSM-V

03. Research Experience

  • Apr 2018 - May 2020 , Tampa, Florida, United States · The University of South Florida

    Oversight in the entire research study process from: the creation of study materials, running participants, wrangling & analyzing data, and synthesizing findings in the form of reports & conferences. Skills and responsibilities include:

    ‣ Generation of standardized reading stimulus representing different experimental conditions for eye-tracking and electroencephalogram (EEG) studies, and integration of said materials to the equipment for administration.

    ‣ Transferring of raw data from Eye-Tracker and EEG equipment to Python, Matlab, and R for data clean-up, statistical analysis (e.g., mixed-effects models), and visualization with ggplot and seaborn.

    ‣ Communicating predictions & results from research studies in the form of conference and grant applications.

  • Feb 2019 - Jun 2019 · 5 mos, Tampa, Florida, United States | The University of South Florida

    Assistance in the administration & preparation of a Depression study for statistical analysis & results delivery

    ‣ Synthesized relevant academic research articles on topics relating to memory & Depression to generate a literature review for the P.hD students of the lab.

    ‣ Encoded data from structured interviews of depressed participants into SPSS for data analysis.Assistance in the administration & preparation of a Depression study for statistical analysis & results delivery