Massry School of Business Abstracts

Presenter(s): Timothy Cleary

Showcase Advisor: Eliot Rich

Abstract: This project examines how a hypothetical New York State agency might use artificial intelligence to reduce office workload, improve processing speed, and rebuild public trust. Using a systems-thinking approach, the project explores how early performance gains from AI may weaken over time if increased reliance leads to staff skill erosion, more mistakes, added cleanup work, and greater oversight demands. The analysis focuses on a two-year rollout period and tracks key variables such as backlog, workload pressure, AI usage, mistakes, cleanup, and trust. Four feedback loops structure the problem: dependence, cleanup, controls, and trust. The project argues that AI adoption is not simply a technical issue, but an organizational one shaped by training, review capacity, and human decision-making. It concludes by identifying practical interventions, including slower rollout, stronger quality checks, better training, and keeping humans in final decision roles.

Presenter(s): Mashaal Munaf

Showcase Advisor: Yeasung Jeong

Abstract: GLP-1 receptor agonists are widely used for the treatment of type 2 diabetes and obesity. Recent research suggests that these medications may also provide additional health benefits beyond weight management, including improvements in cardiovascular, renal, and metabolic outcomes. This study investigates how treatment duration, treatment cost, and side-effect burden are associated with these broader health indicators. Using publicly available health data and clinical research indicators, the study applies descriptive statistics and quantitative analysis to examine relationships between key variables such as A1C levels, systolic blood pressure, BMI, cholesterol levels, and kidney function markers. The analysis aims to identify patterns that may help explain how GLP-1 therapy influences multiple aspects of metabolic health. By integrating healthcare data with analytical methods, this research contributes to a better understanding of the broader effects of GLP-1 therapies and highlights the potential of data-driven approaches in healthcare analytics.

Presenter(s): Leo Meurs

Showcase Advisor: Lee Spitzley

Abstract: Data collection issues have long hindered the study of teams. Significant unsolved challenges include: 1) disruptive measurement practices, 2) overlapping audio/video signals for individuals in a team, 3) accurately synchronized behavioral and activity information, and 4) annotating data quickly and accurately. This project proposes a data collection and processing system to capture team meeting dynamics using an array of microphones and a speech-to-text system.

Presenter(s): Komali Kollapudi, Rachael Oyenola, David Matute-Jimenez

Showcase Advisor: Vinicius Lima

Abstract: Phishing is one of the most common cyber threats organizations face, which is why knowing how to investigate them is a core skill for any security analyst. This project simulates a real-world incident response scenario by analyzing a phishing email sample taken from a public GitHub repository. Using tools such as CyberChef, VirusTotal, etc, we follow the stages of the incident response lifecycle to examine the email for suspicious characteristics, decode hidden or encoded content and verify any suspicious links or domains. By following the incident response lifecycle, we aim to demonstrate how analysts investigate indicators of compromise, analyze phishing techniques and document their findings. The goal is to provide insight into how security teams investigate and respond to phishing campaigns.

Presenter(s): Yujung Hwang, Eric Crespo, Elijah Acero, Raghul Kannappan

Showcase Advisor: Vinicius Lima

Abstract: This project analyzes network traffic to identify malicious activities using two widely adopted network security tools, Zeek and Suricata. The analysis is conducted using the CIC-IDS2017 dataset, a widely used benchmark dataset containing realistic network traffic and multiple modern cyber attack scenarios. In particular, the Wednesday PCAP capture is selected because it includes several attack types within a single dataset, including HTTP-based denial-of-service attacks (Slowloris, Slowhttptest, Hulk, and GoldenEye) as well as the Heartbleed TLS vulnerability attack.

In this project, Zeek is used to extract detailed network activity logs from the PCAP file, including connection, DNS, HTTP, and TLS events. Suricata is then applied to detect potential intrusions using signature-based detection rules. By correlating Suricata alerts with the network logs generated by Zeek, this study aims to identify suspicious patterns and infer attacker behaviors such as denial-of-service activity and abnormal network communication.

Presenter(s): Justin Klotz, Muhammad Bahar, Eleftheria Katsani, Othmane Benkhalifa

Showcase Advisor: Sanjay Goel

Abstract: This presentation will revolve around a Supervisory Control and Data Acquisition (SCADA) environment within a data center, specifically featuring the fire suppression system.

Presenter(s): Emma Feinen, Nicholas Tsilimidos, Amarchi Anyene, anika taylor

Showcase Advisor: Yuksel Celik

Abstract: A miniature SCADA system that simulates an industrial water tank

Presenter(s): Mahar Khan, Theo Rhynie, Emmanuelle Elixir Flores, Alisha Gomez

Showcase Advisor: Vasuda Trehan

Abstract: We will be presenting our app made in Mendix called ScreenLog. A mobile application that serves as a personal media diary. Combining a watchlist with a personal archive, users can discover and review movies in a simple click. Users can track their personal favorites and write notes about what they’ve watched.

Presenter(s): Spencer Kone, Vamshi Krishna Arakala, Anusha Bathini, Chethana Boyapati, Manoj Damor

Showcase Advisor: Sanjay Goel

Abstract: As municipalities transition to smart city architectures, the convergence of electrical grids with IoT-enabled Electric Vehicle (EV) infrastructure creates significant cyber-physical vulnerabilities [cite: 5, 2025-12-19]. This project involves designing and constructing a standalone SCADA training pod that simulates a municipal energy node [cite: 5, 2026-02-26]. The architecture integrates a Field Layer of sensors and actuators with a Control Layer utilizing Programmable Logic Controllers (PLCs) to manage simulated energy loads. The system employs a dual-communication strategy prioritizing Ethernet-based transport, using Modbus TCP for deterministic local control and MQTT for scalable remote telemetry. A primary research objective is to collect and analyze network traffic to identify data anomalies under simulated attack conditions, such as man-in-the-middle and unauthorized command injection. By documenting these exploits, the project aims to develop high-integrity supervision strategies and resilient cryptographic frameworks capable of protecting critical urban infrastructure against emerging cyber threats.

Presenter(s): Trevaughn Little

Showcase Advisor: Heidi Knoblauch

Abstract: In this Case study I will analyze the ways in which modern businesses are adjusting their business strategies to keep up with the competitive demands of a changing business environment. By focusing on the key principles of business administration, such as innovation, leadership, and strategic decision-making, this presentation will discuss the ways in which businesses are adjusting to a changing technological environment. By analyzing trends within the business industry, this project will demonstrate the need for effective business management practices. The purpose of this research is to demonstrate the ways in which effective business strategies are contributing to the growth and sustainability of businesses. This will provide a valuable understanding for future business leaders to navigate a complex world economy.

Presenter(s): Wilson Pauta Jr.

Showcase Advisor: Suraj Commuri

Abstract: Artificial intelligence is reshaping the consumer landscape as we know it. Small businesses and retailers sit at the intersection of America's economic vitality, yet many refuse to adapt to AI-driven tools that are redefining consumer expectations. The literacy gap between large enterprises and small business owners will only continue to grow wider. This presentation encourages business-educated students to give back to their communities by filling that knowledge gap about the evolving consumer market. From this responsibility, we take our first actionable step: awareness. This is the foundation of meaningful economic stewardship. This poster examines how AI is changing consumer behavior, what that means for small retailer survival, and why the next generation of business graduates cannot afford to look away.

Presenter(s): Mahitha Kalinathabotla

Showcase Advisor: Abhishek Ghosh

Abstract: This study analyzes consumer financial complaint data from the Consumer Financial Protection Bureau (CFPB) to examine how complaint characteristics and company response patterns influence complaint resolution outcomes and response times. The CFPB Consumer Complaint Database contains millions of complaints related to financial products such as credit cards, mortgages, and credit reporting services. Using a dataset of approximately 5.6 million resolved complaints, this study applies descriptive statistics and exploratory data analysis to identify patterns in complaint handling and consumer relief outcomes. Preliminary findings indicate that most companies respond to complaints on the same day, although response times are skewed by a small number of extreme delays. Complaints involving more complex financial products also tend to require longer response times.

Presenter(s): Sri Sai Naga Venkata Vara Lakshmi Kanchipati

Showcase Advisor: Abhishek Ghosh

Abstract: Efficient resolution of public service requests is essential for improving citizen satisfaction and urban governance. This study analyzes the time-to-resolution of service requests submitted through the NYC 311 system to understand how operational factors influence response efficiency. The dataset includes approximately 985,000 service requests recorded between November 1, 2025, and January 31, 2026. The initial phase of the research included reviewing existing literature, developing research hypotheses, and conducting descriptive statistical analysis. Summary statistics and visual exploration were used to examine variations in time-to-resolution across complaint types, boroughs, and responsible city agencies. Preliminary results indicate clear variation in resolution times across service categories and geographic locations. The next stage of the study will apply statistical methods, including analysis of variance and regression models, to test whether these differences are statistically significant and to identify patterns that may help improve the efficiency of urban service systems.

Presenter(s): Mariela Santos Reyes

Showcase Advisor: Eliot Rich

Abstract: As space exploration extends deeper into our solar system, the complexity of missions and the volume of generated telemetry data have rendered traditional, rule-based monitoring systems insufficient. This paper examines the critical role of Artificial Intelligence (AI) and Machine Learning (ML) in enhancing spacecraft anomaly detection. Current methodologies rely on predefined settings that lack the adaptability to evolve with new fault patterns, posing a significant risk during unexpected system behaviors—particularly when ground control communication is limited. By shifting toward autonomous, AI-driven models, missions can interpret massive datasets with higher speed and accuracy, identifying both known and unknown anomalies in real time. This work proposes an alternative model using real-time telemetry for unexpected events, providing a transition toward fully autonomous deep-space exploration.

Presenter(s): Christy Zhou, Harshini Ponnam, Naureen Ahmed, Tasneem Al-ghathi

Showcase Advisor: Vasuda Trehan

Abstract: This Mendix application system helps small bakers manage ingredients, calculate the exact cost of each baked item, and determine the profitability of their products. Many bakers struggle to identify the true cost of their recipes and often fail to monitor ingredient quantities accurately, leading to incorrect pricing, reduced profits, and unexpected shortages. The application addresses these challenges by providing an organized platform for ingredient tracking, automatic recipe cost calculation, and profitability analysis. It also includes low-stock alerts to notify bakers when supplies are running low, helping them avoid interruptions in production. By improving inventory control and pricing accuracy, the system enables bakers to make informed business decisions, reduce waste, and maintain consistent product availability. Ultimately, the application supports small baking businesses in becoming more efficient, profitable, and sustainable.

Presenter(s): Gowtham Sundararajan

Showcase Advisor: Yeasung Jeong

Abstract: Most investors evaluate ETFs using total return and volatility but this misses critical information. Two ETFs with identical returns can follow completely different paths: one rises steadily and recovers quickly from losses, while another experiences prolonged drawdowns that test investor patience. Does this difference matter for future performance? 

This study analyzes five major ETFs (SPY, QQQ, IWM, XLF, XLV) over 2023-2025 and reveals that return sequences significantly predict future performance even after controlling for cumulative returns and volatility. ETFs with smoother patterns and faster recovery demonstrate systematically different outcomes. 

Traditional models are leaving money on the table. By ignoring how returns unfold day-by-day, investors and fund managers miss valuable predictive signals. This research provides a more realistic framework for ETF selection, risk management, and portfolio construction—one that aligns with how investors actually experience markets.

Presenter(s): Bharath Chandu Narapinni

Showcase Advisor: Yeasung Jeong

Abstract: Customer reviews are central to online shopping, yet star ratings alone often fail to capture the nuance of customer experiences. This study examines how text‑derived features—sentiment, review length, and helpfulness—complement numeric ratings in shaping review credibility and customer perceptions. Using a cleaned dataset of 1,247 Amazon Electronics reviews, I construct engineered variables including VADER sentiment scores, word‑count‑based review length, and days since release as a lifecycle measure. Descriptive statistics reveal strong positivity bias in ratings, modest average sentiment, and highly skewed distributions for review length and helpfulness. Visual analysis shows that sentiment increases with star rating, longer reviews receive more helpful votes, and sentiment declines over the product lifecycle. These findings demonstrate that textual features provide meaningful explanatory value beyond ratings alone, offering richer insight into how customers evaluate products and how review credibility emerges over time.

Presenter(s): Venuka Satyapriya Addala

Showcase Advisor: Yeasung Jeong

Abstract: This study examines how academic effort, access to technological resources, and student behaviors interact to influence academic performance. While prior research shows that study time, internet access, and alcohol consumption each affect grades, these factors are often analyzed independently. This research takes a more integrated approach by investigating whether study time exhibits diminishing returns and whether its impact depends on contextual factors. Using a dataset of 395 students, the study analyzes the relationship between weekly study time, internet access, weekend alcohol consumption, and final grades. The model includes a squared term for study time to test for nonlinear effects and interaction terms to examine moderating influences. The findings aim to determine whether internet access strengthens the benefits of study effort and whether alcohol consumption weakens them. By analyzing these relationships together, the study provides a more comprehensive understanding of how multiple factors jointly shape student academic performance.

Presenter(s): Swati Keshan

Showcase Advisor: Deborah Snyder

Abstract: Bija AI is an AI-powered venture-readiness concept designed to help student entrepreneurs move from early-stage ideas to funding readiness. Many student founders struggle to organize venture documents, identify gaps in their business plans, and determine when they are ready to pursue investors, grants, or accelerator opportunities. This project explores how AI can provide a guided pathway through document organization, readiness assessment, improvement recommendations, and funding-source matching. The goal of Bija AI is to reduce confusion, improve preparation, and help aspiring student entrepreneurs build stronger, more fundable ventures. This presentation introduces the problem, the proposed platform workflow, its potential value for student founders and entrepreneurship ecosystems, and its relevance to innovation and AI-driven business support.

Presenter(s): Jessica Massas, Derek Valdez, Chloe Heaslip, Tim Wilbur

Showcase Advisor: Vasuda Trehan

Abstract: This application pertains to information systems and business analytics. As a group, we aim to develop a functional application called BookTrax that helps users maintain and achieve their reading goals via an interactive platform. It will consist of features such as a user profile and login, a book tracking meter, reminders and notifications, and a book-rating/review system. This will be complete via utilization of Mendix, an application-creation software. We are employing database structures, business logic and strategy, use-case analysis, the software development lifecycle, documentation, and team communication throughout the progress on this project. Our audience will benefit from a user-friendly, interactive interface that encourages them to keep up with their reading goals on a consistent basis.

Presenter(s): Kayla Strickland

Showcase Advisor: Heidi Knoblauch

Abstract: This case study examines the leadership strategies used by Malachai Hendrix who spoke in our class. The information presented comes directly from interviews and insights shared during the guest speaker session, where we had the opportunity to engage with him. Using case study evaluation methods, I analyzed Malachai’s approach to challenges, decision-making, and innovation within their organization. I also developed targeted questions to gain more focused feedback and better understand the strategies behind their leadership choices. By connecting classroom concepts with a real-world example, this project highlights how leadership skills are applied in practice. For the audience, this analysis offers insight into the practical methods a leader used to navigate challenges, adapt to change, and pursue new opportunities, demonstrating how leadership strategies discussed in theory can translate into real organizational impact.

Presenter(s): Fahim Ahmed

Showcase Advisor: Heidi Knoblauch

Abstract: This presentation provides an overview of the key themes discussed by a guest speaker, whose professional background spans multiple areas of business practice. The session focused on emerging trends, organizational challenges, and the skills required to succeed in today’s competitive landscape. By synthesizing the speaker’s experiences and recommendations, the presentation offers students a clearer understanding of how business concepts translate into real world decision making and long-term career development.

Presenter(s): Matai Hazel

Showcase Advisor: Heidi Knoblauch

Abstract: This case study examines Salah Harris who was recently appointed as youth director for the city of Rensselaer as he works through and understands the public the demand of his new position, he also has a business that has its own separate responsibilities. This study will focus on he manages the demands of his new role while also keeping track of his existing demands. It focuses on leadership choices, communication approaches and organizational skills that will be needed for success in both areas. As a young professional working to support and guide the youth in the city he grew up in, Harris demonstrates leadership and personal ambition at the local level.

Presenter(s): Saanvi Shah, Nithin Sridasyam, Saimanikanta Reddy, Keerthan Teja Ragi

Showcase Advisor: Sanjay Goel

Abstract: Intrusion Detection Systems (IDS) are essential for identifying malicious activities and protecting modern networks from cyber threats. However, traditional IDS methods often struggle to detect complex and evolving attacks. This research investigates the use of Artificial Intelligence (AI) to improve intrusion detection capabilities. The study focuses on the CICIDS2018 Dataset, which contains realistic network traffic including Web attacks, Denial of Service (DoS), and Distributed Denial of Service (DDoS) attacks. Machine learning techniques are applied to analyze network traffic patterns and classify malicious behavior. By evaluating AI-based detection models on this dataset, the project aims to demonstrate how intelligent IDS systems can improve the accuracy and efficiency of detecting modern cyber threats.

Presenter(s): Manasa Reddy Pam

Showcase Advisor: Yeasung Jeong

Abstract: Consumer spending plays a crucial role in driving economic activity, but traditional macroeconomic indicators often pick up on financial stress only after it has already impacted households. This study looks into whether credit-related consumer complaints can act as an early warning sign for changes in consumer spending. By analyzing monthly national data from the Consumer Financial Protection Bureau (CFPB), the Bureau of Labor Statistics, and the U.S. Census Bureau, the research delves into the connections between complaint activity, unemployment, and retail sales growth. It tests if a rise in complaint volume comes before shifts in consumer spending and whether this link becomes more pronounced during times of higher unemployment. By connecting administrative complaint data with macroeconomic indicators, this research aims to uncover if consumer complaint activity can offer valuable insights into emerging financial stress and short-term changes in economic behavior.

Presenter(s): Maryam Barak, Raisa Sara, Ariana Negron, James Sheridan, Chelsea Guyer

Showcase Advisor: Ismet Ozer

Abstract: The daily drive is a to-do list/notes app. This app allows users to create and store various lists. Users are able to create recurring and collaborative lists. You are able to categorize the lists, check tasks off, set reminders for lists, and create sub-tasks. This app will be created through Mendix. We will have a system design, including entities and relationships to demonstrate a functional business design.

Presenter(s): Susritha Vittanala

Showcase Advisor: Yeasung Jeong

Abstract: Heart disease is one of the leading causes of death worldwide, making early identification of risk factors very important. This project examines how different clinical and demographic factors are related to the presence of heart disease using patient health data. The dataset includes variables such as age, gender, chest pain type, blood pressure, cholesterol levels, maximum heart rate, and ST depression during exercise. 
At the current stage, the study focuses on descriptive statistical analysis and visual exploration to better understand the dataset and identify possible patterns. Initial observations suggest that some variables, including cholesterol levels, ST depression, and maximum heart rate, may be associated with heart disease.  

The project is still in progress. Future work will involve applying statistical and machine learning models, such as logistic regression and random forest, to predict heart disease and identify the most significant risk factors.

Presenter(s): Syed Mohammad Ali Kazmi

Showcase Advisor: Sukwoong Choi

Abstract: Municipal 311 systems are largely used to handle non-emergency requests for services. There is limited empirical evidence on the factors affecting the efficient resolution of such requests. In the current study, the authors sought to understand the variations of the resolution of service requests within the 311 system of San Francisco. The authors used the available case records to examine the effects of the type of service, location, and time of submission on the resolution of the requests. The requests for services are categorized into operational types, such as infrastructure services. The spatial and temporal variations of the requests are represented by the police district and calendar effects. The statistical method is applied to identify the patterns of the performance of the services. The results of the study provide useful insights into the operational aspects of the services of the municipality. The results provide useful insights into areas of improvement.

Presenter(s): Steve Correia, Riya Sharma, Afla Rafeek Kalariparambil Abdul Rafeek, Yogeshwar Kotha, Mayur Raj Singh Biasthakur

Showcase Advisor: Sanjay Goel

Abstract: This project explores autoencoder-based dimensionality reduction techniques on SOC-style intrusion detection datasets. Our group will implement multiple autoencoder configurations and use the reduced data to train classifiers, comparing performance with models trained on the original dataset.

Presenter(s): Vaishnavi Mamindla

Showcase Advisor: Abhishek Ghosh

Abstract: This study examines relationship between regional employment fluctuations and discretionary retail spending in the United States using state-level data from the Opportunity Insights Economic Tracker. The main goal of this research is to understand whether changes in labor market conditions can help explain and predict patterns in discretionary consumer spending. Three hypotheses are explored: first, that improvements in employment levels are associated with increases in discretionary retail spending; second, that consumer spending responds asymmetrically to employment shocks, with negative shocks leading to larger reductions in spending than equivalent positive shocks; and third, that the strength of this relationship varies across states. Using weekly employment data and aggregated retail spending measures from 2020 to 2024, the exploratory analysis identifies patterns consistent with these hypotheses. These findings improve our understanding of how labor market conditions influence consumer behavior and may help policymakers and businesses anticipate changes in consumer demand during economic fluctuations.

Presenter(s): Sai Priya Kemburu

Showcase Advisor: Zhuojun Gu

Abstract: Online review platforms generate large amounts of unstructured text that reflect customers’ experiences and opinions. Many platforms summarize these experiences using structured indicators such as star ratings or binary sentiment labels. However, these measures may not fully capture the variation in how satisfaction or dissatisfaction is expressed in written reviews. This study examines whether sentiment probabilities produced by deep learning models can serve as a more refined indicator of customer satisfaction. Using two datasets—IMDB movie reviews and TripAdvisor hotel reviews—the analysis compares model-generated sentiment probabilities with existing satisfaction benchmarks, including binary sentiment labels and 1–5 star ratings. Three recurrent neural network architectures (LSTM, BiLSTM, and GRU) are evaluated to determine how closely their predicted probabilities align with observed satisfaction measures. Statistical analyses, including correlation and regression, are used to assess whether higher satisfaction levels correspond to higher sentiment probabilities and whether variation exists within the same rating categories.

Presenter(s): Venkata Lakshmi Prasanna Gidituri

Showcase Advisor: Yeasung Jeong

Abstract: This study examines how city-level characteristics influence the resolution time of non-emergency service requests reported through 311 systems. Using publicly available 311 datasets from major U.S. cities such as New York City, San Francisco, and Washington DC, the analysis explores how factors including population size, economic conditions, and demographic characteristics are associated with differences in resolution time. Efficient resolution of 311 service requests is important because it reflects how effectively local governments respond to citizen concerns and manage urban services. Through exploratory data analysis, this study identifies patterns in service request resolution across different cities and examines how demographic and economic contexts may affect municipal service delivery. The findings aim to provide insights into how broader city characteristics influence operational efficiency in urban service management.

Presenter(s): Eric Crespo

Showcase Advisor: Prinkle Sharma

Abstract: Connected and Autonomous Vehicles (CAVs) rely heavily on Vehicle-to-Everything (V2X) communication to exchange safety-critical information. However, these networks are vulnerable to Sybil attacks, where a malicious vehicle creates multiple fake identities to manipulate traffic information and disrupt network trust. Traditional rule-based and classical machine learning approaches often struggle to capture the complex relational patterns created by such attacks. This project investigates the use of Graph Neural Networks (GNNs) for detecting Sybil attacks in V2X environments. By modeling vehicles as nodes and communication interactions as edges, graph-based learning methods can capture structural and temporal relationships among entities in the network. The study reviews recent AI-based misbehavior detection approaches and analyzes existing graph-based techniques for Sybil detection. The project aims to identify limitations in current methods, particularly regarding scalability, real-time deployment, and dataset realism, and explores the potential for developing a lightweight and scalable graph-based detection framework suitable for dynamic vehicular environments.

Presenter(s): Silvia Avila Licona

Showcase Advisor: Luis Luna-Reyes

Abstract: While social mobility is often considered a final outcome, it is the result of multiple interconnected social, economic, and institutional factors that have been widely studied in public policy research. This project develops a simulation using a system dynamics model to visualize how public housing policies may influence several key indicators associated with social mobility. These indicators include incarceration rates, educational attainment, employment opportunities, access to community resources, and outcomes for future generations. The model seeks to capture the feedback relationships between housing stability, neighborhood conditions, and long-term opportunities for residents. The project focuses specifically on Albany County, New York, within the context of ongoing political discussions about increasing public investment in housing. By modeling these interactions over time, the study aims to explore how policy interventions in public housing could shape long-term neighborhood dynamics and potentially improve pathways for upward social mobility among low-income populations.

Presenter(s): Sruthi Swaminathan

Showcase Advisor: Yeasung Jeong

Abstract: This research investigates how three host policies—Instant Book, minimum-night requirements, and self check-in—may shape Airbnb performance across New York City, Hawaiʻi, and San Francisco over the last 365 days. Performance is assessed through estimated occupancy and bookings, two indicators that are highly relevant to host revenue and platform competitiveness. This topic is important because host-controlled policies directly influence guest convenience, booking flexibility, and accessibility, all of which can affect traveler decision-making in different market contexts. Understanding these relationships helps explain how operational choices contribute to listing success in urban and tourism-driven destinations. It is also valuable for identifying whether the same strategy works similarly across diverse markets or whether hosts should adapt their policies to local demand conditions. By comparing these three cities, the study provides a practical framework for evaluating host decisions and their potential implications for short-term rental performance.

Presenter(s): Shivshankar Paswan

Showcase Advisor: Abhishek Ghosh

Abstract: Socioeconomic conditions play a critical role in shaping how individuals perceive and experience their health. This study examines the relationship between income, educational attainment, health insurance status, and self-reported general health among adults in California. Using data from the 2024 California Health Interview Survey (n = 24,810), descriptive statistics and visual analysis reveal that higher income and education levels are consistently associated with better self-reported health outcomes. Insured individuals also report slightly better health than their uninsured counterparts. Additionally, income appears to influence the strength of the relationship between education and health, suggesting these factors do not operate independently. These findings highlight how distinct socioeconomic characteristics contribute to measurable health disparities within a large, diverse population, offering meaningful insight for public health policy and targeted intervention efforts.

Presenter(s): Sirisha Usetti

Showcase Advisor: Sukwoong Choi

Abstract: Forecasting stock market volatility is essential for investment decision-making, risk management, and financial stability. Traditional volatility models mainly rely on historical stock price data, which may overlook broader economic conditions that influence financial markets. This study investigates whether incorporating macroeconomic indicators improves the accuracy of stock market volatility forecasting compared to models based solely on historical price data. Using daily S&P 500 index data, monthly volatility is calculated as the standard deviation of daily returns. The analysis includes key macroeconomic indicators such as inflation (Consumer Price Index), interest rates (Federal Funds Rate), unemployment rate, and gross domestic product (GDP). Data are obtained from Yahoo Finance and the Federal Reserve Economic Data (FRED) database for the period 2005–2024. The study compares price-based models with models that integrate macroeconomic variables to evaluate whether economic indicators provide additional predictive power in forecasting fluctuations in stock market volatility.

Presenter(s): Dilruba Koli

Showcase Advisor: Zina Lawrence

Abstract: This project presents an incident response scenario conducted in a controlled virtual environment, employing a Kali Linux system as the attack platform and an Ubuntu host integrated with the Wazuh XDR security monitoring solution. The objective is to simulate a remote access attack against an exposed Secure Shell (SSH) service protected by a weak password. The Kali Linux system initiates repeated authentication attempts to obtain unauthorized access to the target host. Throughout the exercise, Wazuh continuously monitors system events, generates security alerts, and records anomalous authentication activity. These alerts enable the identification of brute-force behavior and potential compromise attempts. The study demonstrates the effectiveness of extended detection and response capabilities in detecting and responding to malicious activity in near real time. Analysis of Wazuh-generated logs and alerts underscores the importance of detection and response solutions for improving visibility, enabling threat detection, and supporting incident response.

Presenter(s): Kaifuddin Azlan Mohammed

Showcase Advisor: Yeasung Jeong

Abstract: This study examines whether Airbnb listings that offer instant booking tend to charge higher prices than listings that require host approval. Instant booking is an important platform feature because it allows guests to confirm reservations immediately, which may make these listings more attractive and allow hosts to charge higher prices. The analysis uses publicly available data from the Inside Airbnb dataset and focuses on listings from three U.S. cities: Austin, Chicago, and Denver. The dataset includes information on listing prices, instant booking availability, review ratings, and the number of reviews. Listing prices are transformed using the natural logarithm to improve reliability of the analysis. A regression model is used to estimate the relationship between instant booking and listing prices while controlling for review ratings, number of reviews, and city differences. The findings provide insight into how platform features and reputation signals influence pricing strategies in Airbnb markets across cities.

Presenter(s): Cyril Thomas, Perry Grandis, Aashish Shrestha, Manoj Damor

Showcase Advisor: Sanjay Goel

Abstract: The rapid expansion of Industrial Internet of Things (IIoT) devices has broadened the cyberattack surface across critical infrastructure, creating an urgent need for intelligent intrusion detection. Traditional signature-based Intrusion Detection Systems (IDS) fail to generalize against novel, multi-vector threats. This paper proposes a deep learning-based IDS to accurately classify and detect cyberattacks in IIoT environments. Deep learning offers key advantages over conventional methods by learning complex patterns directly from raw network traffic while generalizing effectively to previously unseen attack types. The proposed system will be trained and evaluated on the ‘DataSense IIoT Dataset 2025’ (CIC, UNB), which captures 50 distinct attacks across seven categories including DDoS, DoS, Reconnaissance, Web-based, Brute Force, MITM, and Mirai Malware, totaling over 1.8 billion packets collected from 40 interconnected IIoT devices alongside 259,212 benign packets representing normal traffic. Performance will be assessed using accuracy, precision, recall, and F1-score across binary and multi-class classification tasks.

Presenter(s): Michael Lavelle

Showcase Advisor: Heidi Knoblauch

Abstract: This project will traverse the professional journey and leadership style of Ken Kramer, a veteran business leader in the technology and SaaS industries. The case study poster will focus on Kramer's ability to build and manage high-performing teams, organize work efficiently and make strategic business decisions. Utilizing his professional profile and specific examples throughout his career, we will explore how Kramer managed to guide companies through major challenges and upheaving changes like acquisitions, implementation of new systems and other ventures for company expansion. Attendees will learn many lessons on leadership, approaching different challenges in business and transforming an idea into a tangible result; all valuable lessons for normal students and future entrepreneurs alike.

Presenter(s): Selay Eter

Showcase Advisor: Heidi Knoblauch

Abstract: This project examines leadership and entrepreneurship through a case study interview with a business leader. And I picked Gary Goldstein as the leader and entrepreneur. The research explores how the Gary Goldstein identified opportunities, made strategic decisions, and navigated challenges while building and managing their venture. By analyzing their leadership style, risk-taking approach, and impact on their organization or community, the project highlights key lessons about entrepreneurial leadership and innovation in practice.

Presenter(s): Sai deepthi Sapata

Showcase Advisor: Sukwoong Choi

Abstract: This study examines how linguistic characteristics influence reader engagement in human-written books and evaluates whether AI-generated texts exhibit similar linguistic patterns. Using natural language processing techniques, linguistic features such as lexical diversity, sentiment intensity, emotional variation, repetition, and sentence-length variation are extracted from fiction texts obtained from Project Gutenberg and linked with reader engagement metrics from Goodreads. Regression modeling is used to identify linguistic predictors of reader ratings and popularity outcomes. The study then compares these empirically derived patterns with AI-generated narratives produced under controlled conditions. By integrating predictive modeling and comparative analysis, this research contributes to computational text analytics and provides insights into whether AI-generated writing aligns with linguistic characteristics associated with successful human-authored works

Presenter(s): Shriiya Gilla

Showcase Advisor: Sukwoong Choi

Abstract: This study investigates whether textual characteristics of Airbnb listing descriptions and host profile biographies can predict host responsiveness on the platform. Using the Boston dataset from InsideAirbnb, the analysis examines host-written text alongside responsiveness metrics such as response rate and response time. Text mining techniques are used to extract measurable linguistic features, including description length, readability, sentiment, and indicators of professional writing. These features are incorporated into regression and classification models to evaluate their ability to explain and predict host response behavior, while controlling for factors such as host tenure and listing activity. The study aims to determine whether written self-presentation serves as a meaningful signal of operational behavior in peer-to-peer marketplaces. The findings contribute to understanding how text analytics can be applied to identify patterns in host engagement and responsiveness.

Presenter(s): Mikail Demir

Showcase Advisor: Luis Luna-Reyes

Abstract: Court decisions in the United States are technically public, yet practically inaccessible without expensive proprietary systems controlled by a dominant commercial actor. This paper examines the dynamic relationship between legal information access policy and access to justice, using a system dynamics framework. The central argument is that neutral citation adoption — the practice of courts attaching non-proprietary identifiers to their decisions — is the key policy lever that can break incumbent lock-in and lower the cost of legal information access. However, reform is resisted by a reinforcing feedback loop: professional norm entrenchment sustains dependence on proprietary systems, dampening the rate of policy adoption and limiting the growth of competing open platforms. Using Behavior over Time graphs and a causal loop diagram, this proposal maps the hoped and feared trajectories of reform, tracing the chain from legal information accessibility to the cost of legal representation and ultimately to access to justice.

Presenter(s): Eric Crespo

Showcase Advisor: Prinkle Sharma

Abstract: This research project focuses on preparing culturally and demographically diverse datasets for machine learning applications in mental health through robust and interpretable feature engineering. The goal is to create meaningful features that support fair, effective, and generalizable mental health models across diverse populations. Machine learning is increasingly used to detect behavioral patterns, predict mental health conditions, and support clinical decision-making, but model performance depends heavily on the quality and representativeness of the data. A key challenge is harmonizing datasets while preserving important cultural and demographic context. This project follows three phases: first, datasets will be explored and cleaned using statistical analysis and visualization to identify missing values, outliers, and demographic imbalances. Second, domain-informed feature engineering will extract linguistic, behavioral, and psychological indicators while incorporating demographic-aware methods. Finally, model-ready datasets and evaluation scripts will be developed to support fair benchmarking and reproducible research in mental health AI.

Presenter(s): Cory Adams Burke

Showcase Advisor: Luis Luna-Reyes

Abstract: Banks rely heavily on depositor confidence to maintain stable funding. When confidence weakens, withdrawals can accelerate rapidly, creating liquidity stress for otherwise solvent institutions. This project develops a system dynamics approach to explore how depositor confidence, withdrawal behavior, and bank liquidity interact over time. The model represents core banking flows such as deposits, withdrawals, and liquidity buffers, and incorporates feedback loops that capture how declining confidence can amplify withdrawal rates. To improve realism, U.S. bank Call Report data are used to calibrate baseline deposit and liquidity levels. Simulation results illustrate how small shifts in depositor sentiment can propagate through the banking system and potential trigger liquidity crises. The model aims to provide a simplified but informative framework for understanding bank run dynamics and the importance of maintaining depositor trust.

Presenter(s): Nithila Nagamanickam

Showcase Advisor: Yeasung Jeong

Abstract: Consumer complaint systems are widely used to monitor service quality in financial markets, yet the factors that lead consumers to escalate complaints by disputing company responses remain insufficiently understood. This study analyzes a large dataset of credit card complaints to examine the determinants of consumer dispute behavior. Descriptive statistics and visual exploration are used to identify patterns in complaint characteristics, including issue type, submission channel, response timeliness, geographic location, and company. Logistic regression is applied to evaluate how these factors influence the probability of a dispute. In addition, machine learning models, including decision trees and random forests, are used to capture nonlinear relationships and assess variable importance. The findings provide insights into consumer escalation behavior and demonstrate how data analytics can improve understanding of complaint management in financial services.

Presenter(s): Ethan Maliszewski

Showcase Advisor: Luis Luna-Reyes

Abstract: Since the 2000s, the sale and use of fentanyl have rapidly grown within the United States illicit drug market, contributing to the public health crisis of overdose deaths. Based on research by Caulkins, Bushway, Milward, and Reuter, traditional drug policy approaches applied to illicit drugs may not effectively reduce fentanyl supply and trafficking across the border. Fentanyl requires less physical space for transport across the border than other illicit drugs while maintaining its potency and addictive properties. Producers have shifted to converting fentanyl powder into counterfeit prescription pills to increase profits and concealment. Due to these distinctive properties, drug enforcement may have a limited effect on reducing distribution and use and can unintentionally contribute to inter-cartel violence. This model examines the effects of drug enforcement on the supply and demand of fentanyl within the United States. Market adaptation is explored through balancing and reinforcing feedback loops.

Presenter(s): Christopher Murunguzha

Showcase Advisor: Luis Luna-Reyes

Abstract: Auto insurance fraud remains a persistent and costly challenge within New York State’s no‑fault insurance system, driving up claim costs and contributing to rising premiums for policyholders (Linders, 2026; Hochul State of the State Address 2026). This study employs a system dynamics approach to investigate the structural drivers and behavioral patterns that enable fraud to grow and persist within the state’s auto insurance industry. The analysis focuses on how interactions among claimants, medical providers, attorneys, insurers, and regulatory agencies generate feedback mechanisms that either reinforce or counteract fraudulent activity. Using a stock‑and‑flow modeling framework, the study represents the accumulation of fraudulent claims, the emergence of organized fraud networks in the form of stage crashes, and the influence of investigative and regulatory interventions.

Presenter(s): Claudia Dispinzeri

Showcase Advisor: Zhuojun Gu

Abstract: This study investigates whether mortgage approval decisions remain consistent across financial institutions when applications share comparable economic characteristics. Using large-scale transaction-level data from the U.S. Home Mortgage Disclosure Act (HMDA) covering the period from 2018 to 2024, the project examines how institutional identity relates to approval outcomes after accounting for borrower income, loan amount, state, product type, and year. The analysis focuses on final underwriting decisions (approved or denied) and evaluates whether systematic variation persists across institutions operating under the same regulatory environment. By combining exploratory data analysis with statistical and machine learning models to estimate approval probabilities, the study assesses whether institutional identity provides additional explanatory power beyond observable economic characteristics. The results contribute to understanding institutional heterogeneity in mortgage lending and demonstrate how large administrative datasets can be used to audit decision consistency in regulated credit markets.

Presenter(s): Somya Luthra

Showcase Advisor: Luis Luna-Reyes

Abstract: The rise of Artificial Intelligence in K-12 education has intensified calls for teacher professional development (PD), yet significant gaps remain in training access and outcomes. While federal policy has prioritized AI competency among students, research shows that teachers, particularly in high-poverty districts, are underprepared to meet this demand. Existing studies capture static outcomes, failing to account for dynamic, interconnected processes through which policy decisions, training investments, teacher perspectives, and student learning outcomes mutually reinforce or undermine one another over time. This study proposes the first system dynamics model of AI competency in K-12 education, focusing specifically on teacher PD. By mapping feedback loops among variables such as teacher AI literacy, competency, workload, burnout, institutional capacity, and student outcomes, the model identifies key policy intervention points and surfaces unintended consequences that conventional research methods cannot detect. Findings offer actionable guidance for educators, administrators, and policymakers designing equitable, sustainable AI-competency PD programs.

Presenter(s): Kardelen Mirlay

Showcase Advisor: Sukwoong Choi

Abstract: Marketing campaigns are expensive endeavors for any organization. Organizations are increasingly turning to customer analytics to optimize the marketing campaigns they have. This project seeks to determine the factors that are associated with customer response to marketing campaigns. This is done by integrating the following mechanisms from database marketing: customer capacity/heterogeneity, customer-firm relationship strength, and opportunity for offer exposure. The research will be based on the following data from the Kaggle Customer Personality Analysis data set: how the characteristics of the customer, such as income level and age, relate to the customer response to the marketing campaign. The research will be to create a research model based on the established marketing theories and strategies with a clear definition of the variables involved.

Presenter(s): Madhu Chockalingam

Showcase Advisor: Yeasung Jeong

Abstract: This study examines how inflation, wage growth, and consumer sentiment are related to labor force participation in the United States and how these relationships changed before and after the COVID-19 pandemic for prime age workers. Understanding these relationships is important because economic conditions influence individuals’ decisions to enter or remain in the labor force, which affects overall economic stability and growth. The analysis will use publicly available U.S. macroeconomic data, including labor force participation rates, inflation measures, wage indicators, and consumer sentiment indexes. Using descriptive and comparative analysis, the study will explore trends and correlations across the pre-COVID and post-COVID periods to identify potential changes in labor market dynamics. The findings aim to provide insights into how economic conditions shape labor supply behavior and contribute to discussions on post-pandemic labor market recovery.

Presenter(s): Prajkta Patil

Showcase Advisor: Zhuojun Gu

Abstract: This study examines how financial indicators can be used to understand and predict revenue growth in AI-enabled firms. As artificial intelligence continues to shape business strategy and operations, many firms are investing in AI technologies to improve efficiency, innovation, and decision-making. However, financial outcomes vary widely across companies, making it important to identify the conditions associated with stronger revenue performance. Using publicly available firm-level financial data, this research explores the relationship between revenue growth and key indicators such as profitability, liquidity, leverage, firm size, and net profit margin. The study applies descriptive analytics to identify patterns in financial performance, predictive analytics to estimate revenue growth, and prescriptive insights to highlight conditions linked to stronger outcomes. By combining these approaches, the research aims to provide data-driven guidance for understanding growth patterns and supporting strategic decision-making in technology and AI-focused firms.

Presenter(s): Guadalupe Vargas

Showcase Advisor: Heidi Knoblauch

Abstract: This case study examines Richard Lin, an entrepreneur and media professional in New York’s Capital Region who uses digital media to support innovation and entrepreneurship. Lin graduated from Rensselaer Polytechnic Institute, where he studied computer science and business while becoming involved in the region’s startup community. During his time at RPI, he worked with entrepreneurial programs and events that connected students, founders, and innovators. Through these experiences, he recognized a need for high quality media production and livestreaming services that could help organizations share ideas and reach audiences. In response, he founded Agora Media, a company dedicated to producing video content, livestreaming conferences, and supporting events across the region. Through Agora Media, Lin helps startups, universities, and organizations communicate their work effectively. His career shows that media and communication are often treated as secondary to innovation, yet they play a role in helping ideas spread, connect, and come to life.

Presenter(s): Natalie Metz

Showcase Advisor: Luis Luna-Reyes

Abstract: Hepatitis C Virus (HCV) remains a major public health concern in the United States, particularly among people who use drugs (PWUD). In 2021, New York State (NYS) announced a statewide plan to eliminate HCV by 2030. Given the high concentration of new HCV cases, the plan focused primarily on injection drug use. However, as smoking has become a more common modality of drug use, understanding how people access and share smoking supplies, as well as their knowledge about HCV is increasingly important for reducing HCV risk. We completed in-depth interviews with 49 PWUD about substance use patterns and access to smoking supplies. Using a system dynamics perspective, explores feedback loops between clients, programs, and community support. This work highlights the feasibility and acceptability of an innovative harm reduction strategy that may help NYS achieve HCV elimination.

Presenter(s): Shreyas Mysore Sudesh

Showcase Advisor: Yeasung Jeong

Abstract: Cities generate continuous records of public service interactions, yet whether response times reflect operational capacity or structural inequality remains empirically underexplored. This study constructs a real-time predictive pipeline on Google Cloud Platform to continuously ingest live NYC 311 service requests via the SODA API and forecast resolution latency at the point of ticket submission. A Medallion architecture streams raw events into Cloud Storage, enriches them with neighborhood-level socioeconomic indicators through BigQuery spatial joins, and trains an XGBoost gradient-boosted regressor using BigQuery ML. Exploratory analysis of over 3.4 million records reveals that lower-income neighborhoods exhibit systematically higher resolution latency: a gap that persists across agencies, complaint types, and boroughs. The model's feature importance output quantifies whether neighborhood income independently predicts slower service after controlling for agency workload, transforming observed disparity into measurable, evidence-based grounds for municipal accountability.

Presenter(s): Maureen Mbanga

Showcase Advisor: Guy Fernando

Abstract: Artificial Intelligence (AI) is rapidly transforming the fields of accounting and finance. This capstone examines the growing role of AI in modern accounting, focusing on how it automates routine tasks, analyzes large financial datasets, and supports more accurate decision-making. The project explores key questions students’ perception of how AI will affect the future of the profession. To investigate this topic, the study uses academic research on AI technologies in accounting along with qualitative data collected through interviews with accounting students about their perceptions of AI and its potential impact on employment in the profession. Understanding these developments is important for students, educators, and professionals because AI is reshaping required skills and career expectations in accounting. While AI can increase efficiency and insight, the research highlights that professional judgment, ethics, and human oversight remain essential to the accounting profession.

Presenter(s): Jack Reece

Showcase Advisor: Heidi Knoblauch

Abstract: The Standard Hydrogen Corporation revolutionizes the way we store and distribute renewable power and fuel. They are building a network of multi-functional, onsite produced renewable hydrogen stations to deliver needed benefits to NY's rapidly expanding renewable energy market, and to energize the zero emissions vehicles (ZEVs) of tomorrow. As part of BMGT480, the co-founder of SHC, William Dailey, came to speak about his career path, where we were able to connect during his visit. This project will be a case study on how Standard Hydrogen Corporation relies on William Dailey to lead their advancement of renewable energy in the context of New York State's push for a clean energy economy.

Presenter(s): Thanh Ngo

Showcase Advisor: Luis Luna-Reyes

Abstract: This project investigates the systemic factors contributing to the housing affordability crisis in New York City. This analysis focuses on variables including land availability, housing supply, population dynamics, and economic conditions such as cost of living and employment. The analysis explores how the interactions of these variables shape housing demand, influence price trajectories, and ultimately determine affordable housing accessibility for residents. This study also examines the historical behavior and reference modes of these variables to contextualize long-term trends. Employing a systems thinking framework, the project emphasizes identifying underlying patterns and feedback structures within the housing system rather than forecasting a single deterministic outcome. Through system dynamics modeling, the research simulates a range of policy interventions to evaluate how modifications in the aforementioned variables could affect housing affordability over time. The goal is to generate insights that support more effective policy design aimed at alleviating New York City’s escalating affordability challenges.

Presenter(s): Malena Spiesz-Aughtry

Showcase Advisor: Heidi Knoblauch

Abstract: This case study explores the professional journey of Chuck Reed, founder and CEO of CBR Improvement Strategies. After building a successful career in corporate leadership, Reed made the decision to transition into entrepreneurship and establish his own consulting firm focused on organizational improvement and leadership development. This study will examine the motivations behind his decision to leave the corporate world, the challenges associated with launching a consulting business, and the leadership philosophy that guides his work today. This presentation will explore themes of strategic decision making, leadership, and entrepreneurship while encouraging discussion about risks and opportunities associated with pursuing an independent career path.

Presenter(s): Ryan Keating, Joshua Allen

Showcase Advisor: David Smith

Abstract: UASBIG provides an environment conducive to the education of undergraduate students who have a genuine interest in and passion for the field of finance. The group affords students the opportunity to learn proper research techniques for successful portfolio and investment management and to manage a portfolio of live assets.

The goal of the UASBIG common stock portfolio is to outperform the returns of the S&P 1500 Index by maximizing the value of assets under management with adherence to a sound long-term investment policy.

We will be delivering 2 pitches to simulate the process of modeling a company and how it would get into the portfolio. Advisory Board members would be in attendance asking us questions regarding the company and the assumptions we made in our model.

Presenter(s): Shalem Raju Maddirala, Achyunt Lamichhane, Perry M Grandis, Atharva Shigvan, Junaid Mohammad

Showcase Advisor: Sanjay Goel

Abstract: Digital gaming and esports environments provide structured, goal-oriented platforms that support learning, social interaction, and cognitive development. For neurodiverse individuals, structured esports programs offer predictable environments that encourage teamwork, communication, and engagement. This research explores how university esports programs can support social and cognitive outcomes. This study investigates how structured esports environments in universities can promote social engagement and cognitive development for neurodiverse participants.

Presenter(s): Amber Lucky

Showcase Advisor: Eliot Rich

Abstract: Over the past two decades, state and local governments have increasingly relied on tax incentives to attract firms and stimulate economic growth. While these policies are intended to generate jobs and investment, research suggests their economic benefits are often limited. This project examines corporate tax incentives as a dynamic policy problem shaped by interstate competition, political incentives, and fiscal tradeoffs. Drawing on system dynamics, the analysis explores how reinforcing competitive pressures and short-term political rewards encourage the continued use of incentives, even when their long-term economic impacts remain uncertain. At the same time, reduced tax revenues may constrain governments’ ability to fund public goods that support sustainable growth. Understanding these feedback processes helps explain why incentive policies persist despite mixed evidence of their effectiveness. The project develops a conceptual system dynamics framework that maps these feedback processes and provides a foundation for future quantitative modeling of state tax incentive competition.