Chelsea Anestal | MBA Student Portfolio

About Me

I am a technical consultant and MS/MBA student focused on the intersection of finance, analytics, and responsible technology. My work combines quantitative modeling, business strategy, and applied AI to evaluate risk, improve decision quality, and deliver practical solutions across healthcare and financial contexts.

Skills

Python Flask Django FastAPI Docker AWS GCP Azure SQL NoSQL PySpark Tableau PowerBI R Machine Learning Artificial Intelligence Data Visualization Financial Risk Management AI Governance Regulatory Compliance (GDPR, CCPA, HIPAA) Operational Risk Assessment Cost Mitigation Planning Enterprise Risk Strategy

Financial Modeling & Analytics Portfolio

This portfolio section highlights applied MBA finance and analytics coursework focused on forecasting, investment analysis, portfolio construction, and financial automation.

1. Forecasting & Time Series Analysis

Module: Mod 1 - Forecast Time Series

Built forecasting models using historical data to identify trends, evaluate patterns, and support forward-looking business and financial decisions.

Key Skills: Time series analysis, forecasting, trend analysis, Excel/statistical modeling, business interpretation.

Download Module File

2. Dataset Analysis & Business Insights

Module: Mod 2 - Analyzing Datasets

Analyzed structured datasets to identify patterns, summarize key findings, and translate data into actionable business insights.

Key Skills: Data cleaning, exploratory data analysis, descriptive statistics, business analytics, visualization.

Download Module File

3. Capital Budgeting & Risk Analysis

Module: Mod 3 - Capital Budgeting and Risk Analysis

Evaluated investment decisions using capital budgeting methods and risk analysis to assess project feasibility and expected financial value.

Key Skills: NPV, IRR, payback period, sensitivity analysis, scenario analysis, risk-adjusted decision-making.

Download Module File

4. Portfolio Analysis

Module: Mod 4 - Portfolio

Developed a portfolio analysis model to evaluate investment performance, diversification, risk-return tradeoffs, and asset allocation strategies.

Key Skills: Expected return, volatility, covariance/correlation, diversification, efficient frontier, capital market line, portfolio optimization.

Download Module File

5. VBA Financial Automation

Module: Mod 5 - VBA

Used VBA to automate financial modeling tasks, improve spreadsheet functionality, and streamline repeatable analytical workflows.

Key Skills: Excel VBA, automation, macros, spreadsheet modeling, process efficiency.

Download Module File

6. Options & Risk Management

Module: Mod 6 - Options

Analyzed options strategies, including protective puts, option moneyness, and delta, to evaluate downside protection and risk management opportunities.

Key Skills: Options analysis, protective puts, delta, moneyness, derivatives, downside risk management.

Download Module File

7. Final Project: Investment & Financial Modeling Analysis

Title: Financial Modeling & Investment Analytics: Multi-Sector Equity, Options, and Portfolio Analysis

Completed a comprehensive financial modeling project analyzing selected equities across multiple industries, including technology, finance, and healthcare. The project incorporated return analysis, risk assessment, options strategy evaluation, and portfolio optimization to assess investment performance and downside protection.

Analyzed multi-sector equities using historical market data to evaluate return, volatility, and comparative investment performance.
Assessed portfolio risk-return tradeoffs using diversification concepts, efficient frontier analysis, and capital market line visualization.
Evaluated options-based risk management strategies, including protective puts, moneyness, and delta sensitivity.
Applied financial modeling techniques to support investment decision-making and communicate findings through structured analysis.

Download Final Project File

Financial Performance Analysis: Efficiency, Leverage, and Firm Performance

Analyzed firm-level financial performance across 100 companies over a 10-year period to evaluate the role of leverage, efficiency, revenue momentum, and profitability. Used Excel, R, and Tableau to perform exploratory analysis, build visualizations, and communicate insights on financial drivers.

Analyzed 1,000 firm-year observations across revenue, profitability, leverage, efficiency, cash flow, and engineered time-series variables.
Used Excel for descriptive statistics and pivot-table analysis, R for distribution and relationship analysis, and Tableau for dashboard visualization.
Identified stable profitability patterns across firms, with average net margin around 14%, EBIT margin around 19%, and gross margin around 42-43% over the 2016-2025 period.
Found that leverage and efficiency had weak standalone relationships with profitability, suggesting financial performance was more multifactorial than driven by a single ratio.
Built visual analyses including histograms, boxplots, scatter plots, correlation heatmaps, revenue momentum heatmaps, treemaps, and Tableau dashboards.

Skills: Excel, R, Tableau, financial analysis, data visualization, exploratory data analysis, correlation analysis, dashboarding, financial ratios, business storytelling.

Project Files:

Download Project Report (DOCX) Download Excel Numerical Summaries Download Excel Graphical Visualizations Download Tableau Dashboard Workbook Download Tableau Graphical Workbook

FeaturedBusiness Predictive Analytics: Credit Scoring, Churn Prediction, and Cost Optimization

Completed a multi-part business analytics project applying predictive modeling, machine learning, and optimization techniques to solve business problems across credit risk, customer retention, and supply chain cost minimization. Compared regression, classification, and optimization approaches using model performance metrics, cross-validation, ROC-AUC, lift curves, and decision-variable analysis.

Built and compared supervised regression models to predict customer credit scores using financial and behavioral variables.
Evaluated Regression Tree, Random Forest, k-NN, Neural Network, and Regularized Regression models using RMSE, MAE, R-squared, train-test partitioning, and 10-fold cross-validation.
Developed classification models to predict telecom customer churn and evaluated performance using ROC curves, AUC, and lift curves.
Identified Random Forest Classification as the strongest churn model, achieving an AUC of 0.9029.
Created an optimization case study for Walsh's Juice Company to minimize transportation and processing costs using shipment and production decision variables.
Translated technical outputs into business recommendations across credit risk, customer retention, and operational cost management.

Download Project 2 Analytics Report (DOCX)

Selected Technical & Research Projects

FeaturedEnterprise AI Decision-Support System for Student Retention and Advisor Workflow Optimization

Project Summary: Designed an AI-powered Student Success Early Warning System for a mid-size university to identify at-risk first- and second-year students earlier, improve advisor prioritization, and support retention. The proposed system integrates Student Information System, Learning Management System, and cafeteria POS data into advisor-facing dashboards, using machine learning risk scoring, explainable AI, MLOps, and governance controls to support timely, ethical, human-centered intervention.

Designed an enterprise AI decision-support concept to help universities identify at-risk students earlier and improve retention outcomes.
Proposed integration of SIS, LMS, and cafeteria POS data to generate student risk scores and advisor-facing alerts within existing dashboards.
Defined measurable success metrics, including reducing first-year dropout/non-return rates by 2-3 percentage points, reducing failing grades by 5 percentage points, and increasing early risk detection by 30-40%.
Recommended supervised machine learning using Gradient Boosted Trees/XGBoost for tabular student data due to interpretability and advisor trust needs.
Designed governance considerations around fairness, differential privacy, explainable AI, bias auditing, and responsible use of behavioral data.
Proposed a human-centered adoption strategy that positions AI as an advisor "co-pilot" rather than a replacement for human judgment.
Developed future-state recommendations for MLOps, model monitoring, personalized intervention recommendations, advisor training, and long-term student success intelligence.

Skills: AI strategy, enterprise AI, decision support, machine learning, XGBoost, MLOps, explainable AI, responsible AI, data governance, workflow optimization, higher education analytics, predictive analytics, stakeholder adoption, dashboard integration, process transformation.

Download Group 9 Final Project Report (DOCX)

Accounting Assistant Chatbot for Financial Question Support and Workflow Guidance

Project Summary: Developed an accounting-focused chatbot designed to help users answer accounting questions, navigate common financial concepts, and support basic accounting workflows. The project applied application development concepts to a finance use case, demonstrating the ability to translate business requirements into a functional technology solution.

Designed and developed an accounting chatbot to support users with accounting-related questions and financial concept guidance.
Applied application development principles to translate a finance/business problem into a functional user-facing tool.
Built a conversational interface to improve accessibility of accounting support and reduce friction for users seeking quick explanations.
Structured the bot around common accounting concepts, workflows, and decision-support needs.
Demonstrated the intersection of finance, AI, and application development through a practical business technology solution.

Skills: Application development, chatbot design, conversational AI, accounting workflows, financial literacy, business requirements, user experience, process automation, AI-enabled decision support, Python, JavaScript, React, Node.js, SQL, APIs, prompt engineering, OpenAI API, database design, front-end development, back-end development.

Open Accounting Assistant Chatbot Download Chatbot File (HTML)

Healthcare Cybersecurity Compliance: HIPAA, HITECH, & Financial Risk

Date: May 2025

This graduate-level research project explores the intersection of healthcare cybersecurity, regulatory compliance, and financial risk. The paper examines HIPAA and HITECH frameworks, real-world cybersecurity incidents like the Change Healthcare breach, and vendor accountability concerns. It also proposes forward-looking policy and technical recommendations.

Incorporated academic, regulatory, and industry sources (CDC, HHS, IBM, etc.)
Highlighted issues with de-identified data, ethical AI use, and third-party risk
Earned an A+ and described by faculty as “publication-ready”

View Full Paper (PDF)

AI-Powered FinTech App

Background: Managing finances—both personal and professional—can be complex. This project focuses on building a secure, AI-powered fintech application that offers insights, suggestions, and tools to improve financial decision-making.

Objective: Design and develop a scalable, secure mobile application for financial data management with AI-enhanced features.

Tools & Technologies:

Frontend: React Native (cross-platform mobile)
Backend: Node.js / Django / FastAPI
Database: PostgreSQL (hosted on Render)
Auth & KYC: Stripe Identity
APIs: Open Banking APIs
Hosting: Railway
Security: SSL, encrypted DB fields, 2FA (Firebase)

Process & Contribution:

Engineered a secure architecture for handling sensitive financial data
Integrated real-time banking APIs for live financial insights
Implemented KYC and 2FA for enhanced user security
Handled full-stack development from UI/UX to deployment

Outcome: Delivered a robust, scalable fintech solution that enables users to manage and analyze financial data securely.

Reflection: This project sharpened my full-stack development skills, particularly in secure data handling and integrating AI-driven features for real-world financial use cases.

EHR Adoption DataSet Analysis – Case Study

Background: EHRs are foundational to modern healthcare, but adoption varies across institutions. This project explored adoption trends in non-federal acute care hospitals across the U.S.

Objective: Clean, model, and visualize a large-scale dataset to identify patterns in EHR adoption and usage, supporting data-driven healthcare strategy.

Tools & Technologies: PySpark, Pandas, NumPy, RandomForestRegressor, Big Data processing, Jupyter/Databricks, Data.gov dataset.

Process & Contribution:

Imported and validated raw healthcare data from Data.gov
Renamed schema fields for clarity and dropped non-essential columns
Handled missing values using a multimodel approach: RandomForestRegressor for high-value features and group-based or global mean imputation for the rest
Engineered a cleaned dataset ready for advanced visualization and analysis

Outcome: A structured and enriched dataset showcasing hospital EHR trends, enabling strategic insights for research and policy work.

Reflection: This project strengthened my ability to work with large healthcare datasets and build pipelines that integrate machine learning with meaningful real-world outcomes.

View Project on GitHub

View EHR Dashboard

The Digital Transformation of Healthcare

Project Type: Individual Academic Research

Tools and Technologies:Tableau, Excel, FRED, Data.gov, HealthIT.gov, Census, World Bank

Summary: This solo project explores how digital technology and health IT adoption shape workforce dynamics, facility readiness, and patient access. I analyzed five public datasets and built dashboards to identify regional and structural disparities in EHR use and digital engagement.

Key Highlights:

Created interactive dashboards visualizing: EHR adoption by state and facility type, HIT employment and healthcare expenditure trends, and Relationship between household tech access and digital care uptake.
Delivered insights on how digital infrastructure affects health equity
Proposed research directions on patient outcomes and health IT gaps

Watch the Presentation View Dashboards

AI Governance and Financial Risk Mitigation for a Global ERP Initiative

Role: Project Lead

Type: Group Strategy & Governance Plan

Course: ISM 6155 – Enterprise Information Systems Management

Challenge

Emerald Hotels & Resorts planned to deploy an AI-enhanced Oracle ERP system globally. As project lead, I oversaw the development of an AI governance framework and comprehensive risk mitigation plan to ensure responsible use of AI across departments and regions.

Key Finance-Focused Contributions

Financial Risk Identification: Mapped operational risks to financial exposure, using impact/likelihood models to prioritize risk controls
Compliance Alignment: Integrated global data privacy laws into governance practices to mitigate multi-million dollar penalty risks.
Governance Structuring: Assigned financial accountability roles across Compliance, Data Privacy, and IT units.
Continuity Planning: Developed a risk-resilient implementation timeline to prevent costly downtime during ERP system cutovers.
Training Programs: Designed employee modules to reduce financial risks tied to AI errors, data leaks, and compliance breaches.

Outcome

Delivered a financially resilient AI governance and risk management framework ready to scale internationally. Structured the plan to safeguard revenue continuity, enhance investor confidence, and reduce total cost of compliance over time.

Core Skills Demonstrated

View Report

Independent Study: AI Platform Development

Developed and maintained server-based AI platforms for research support. Handled data analysis, visualization, and presentations to guide data-driven decision making.

Tech: Python, Docker, Flask, Django, FastAPI, JavaScript, TypeScript, SQL, NoSQL, AWS, GCP, Azure

AI-ML-Exploratory Work

Explorations in decision trees, ensemble methods, and neural networks.

View Project on GitHub

Research Chattr Platform

Built a customizable AI chatbot platform to support research teams in interactive studies.

View Project on GitHub

ResearchChat Platform

A live webpage enabling human-AI interaction for behavioral research and conversational analysis.

Visit Repository

SWARM Algorithm Optimization

Designed a flexible optimization framework with custom constraints and topologies for domain-specific solutions like EHR imputation.

View Project on GitHub