Portfolio

Juliania
Shyprykevych

Technical Experience

Undergraduate Researcher

Dr. Siddhartha Roy's Lab, Rutgers University

May 2026 — Present

Building a mobile app that uses smartphone images to estimate chlorine concentration in water samples in real-world testing scenarios
Diagnosed accuracy drop-off in existing regression models across concentration ranges, driving ongoing model improvements
Comparing regression (ElasticNet, Lasso) and CNN-based models, visualizing accuracy and overfitting tradeoffs under mobile deployment constraints
Collected 342 smartphone images across varying concentrations and lighting to expand and diversify the training dataset

Python PyTorch NumPy Pandas

Siddhartha Roy's Group at Rutgers University

Autograder Developer | Data Structures Lead TA

Rutgers Department of Computer Science

Jan 2026 — Present

Writing test cases and autograders to verify 10,000+ student submissions per semester across 750+ students
Releasing ~12 labs and 3 major coding assignments using Carnegie Mellon's Autolab platform
Holding weekly office hours to support student learning

Java Autolab Makefile

Fellow

Break Through Tech AI Program

May 2026 — Present

Selected from 3,000+ applicants for Cornell Tech's Break Through Tech AI Program
Built foundational ML/AI skills through an online certification from Cornell University

Python Pandas Numpy Scikit-Learn Tensorflow MatPlotlib Seaborn OpenCV

Java Developer & Mentor

Assignment Guru

Jul–Aug 2025 · Jun–Jul 2026

Designed and developed Java programming assignments for 1,000+ Rutgers CS students, highlighting real-world applications of computing
Built sample answer keys and test cases in collaboration with peers and mentors
Illustrated 100+ pixel art graphics to boost assignment engagement

Java Java Swing

Research

Low-Cost AI/ML Water Quality Monitoring

Rutgers RAD Collaboratory (May—Jul 2026) → School of Environmental and Biological Sciences, under Dr. Siddhartha Roy (Aug 2026—Present) Role: Undergraduate Researcher

Overview

Spectrophotometers are a widely-used tool for detecting contaminants in water (such as chlorine) but they're expensive and impractical for field use, especially for rural communities who could benefit from water quality testing. Too little chlorine in water can leave it unsafe due to a lack of protection againt bacteria, while too much poses health risks to people who consume it. My research explores whether computer vision and machine learning can replace them: training models to read contaminant concentrations directly from photos, cheaply and accurately enough to work outside a lab. I joined this project in May 2026 through Rutgers' Chancellor's Signature Initiatives Summer Undergraduate Research Fellowship, a competitive program pairing undergrads with faculty-led research. When the fellowship wrapped in July, my PI gave me the opportunity to stay on, and I've continued the work since. The work is now conducted within School of Environmental and Biological Sciences rather than New Brunswick Office for Research, where the fellowship was hosted.

What I've Built

Trained and iterated on CV/ML models that estimate contaminant concentration from images, aiming to match spectrophotometer-level accuracy at a fraction of the cost.
Generated custom training datasets from lab experiments, controlling for lighting, angle, and sample conditions to keep the data usable for real-world deployment, not just controlled settings.
Optimized image-based quantification algorithms in Python/scikit-learn, testing preprocessing and feature-extraction approaches to improve model reliability across varied lighting and camera conditions.

Current Focus

Improving model accuracy through iterative training/testing across different dataset compositions.
Reading and applying findings from published literature to refine our approach.
Building and hardening the mobile app for real field deployment. I'm working on moving it from a research prototype toward something usable outside the lab.

Research poster for Low-Cost AI/ML Water Quality Monitoring — Research Poster, presented at the CSI SURF Symposium on July 31st, 2026.

Leadership & Involvement

Data Structures Lead Teaching Assistant — Autolab

Rutgers Department of Computer Science

Jan 2026 — Present

One of three Lead TAs for a 750+ student course, specifically responsible for the Autolab grading infrastructure.
Build the autograders for each assignment and release ~12 labs and 3 major assignments per semester on Autolab.
Hold weekly office hours, helping students debug code, work through course concepts, and clarify assignment requirements.

Code Coach

theCoderSchool Montgomery

Feb 2025 — Present

Working privately with students of varying skills and ages (elementary, middle, and high schoolers) to create coding projects.
Teaching coding concepts to advance students' learning based on their needs and interests and providing a personalized and unique experience for each student.
Ensuring each student has a project completed within four months and writing notes and summaries of each lesson to be sent to the parents.
Working in a 2:1 and 1:1 teaching ratio with an individualized lesson for each student.
Communicating with parents after sessions to share updates on their child's progress.

Computer Science Learning Assistant

Rutgers Learning Centers

Aug 2025 — Dec 2025

Ran 100+ minutes of weekly group sessions for CS111 (Introduction to CS) students, reinforcing core programming concepts outside of lecture.
Guided college-level students to gain proficiency in Java, gain debugging and analyzing skills, and practice writing programs and algorithms.

Den Leader (Hnizdova)

Plast Ukrainian Scouting Organization

Sep 2025 — Present

Lead weekly educational and recreational activities for a group of children ages 6—11, focusing on leadership, teamwork, and cultural education.
Manage parent authorization paperwork and other program logistics to keep the group compliant with organizational requirements.
Help coordinate larger, organization-wide events spanning multiple dens.

Teaching Assistant (Game Development)

Girls Who Code

Jun 2025 — Aug 2025

Led a 15-student advisory group, building a supportive community for underrepresented students in tech.
Co-ran daily p5.js sessions for a 70-student cohort as part of a national teaching team.
Held regular student hours to debug code and support project progress.

Technical Experience

Projects

Research

Low-Cost AI/ML Water Quality Monitoring

Rutgers RAD Collaboratory (May—Jul 2026) → School of Environmental and Biological Sciences, under Dr. Siddhartha Roy (Aug 2026—Present) Role: Undergraduate Researcher

Overview

What I've Built

Current Focus

Skills

Programming Languages

Machine Learning & Data

Web & UI

Tools & Platforms

Leadership & Involvement

About

I'm a Computer Science and Cognitive Science student at Rutgers University with a focus on machine learning and applied software development.

I'm currently looking for software engineering and machine learning internship opportunities for summer 2027. Feel free to reach out to me at juliania.shyprykevych@rutgers.edu or explore my work below.

Bachelor of Science in Computer Science; Bachelor of Arts in Cognitive Science | Expected Graduation May 2028 | 3.6 GPA | New York City Metropolitan Area