About Services Client Outcomes Tools Optimizer Contact
Financial Analysis · Data Science · Consulting

Turning complex data
into clear decisions.

Financial Analyst · Data Scientist · Consultant

I work at the intersection of finance, strategy, and AI — building models and tools that drive smarter business outcomes.

Explore My Work 📅 Schedule a Call Get in Touch
About Me

Bio & Background

Cesar A. Gonzalez
Cesar A. Gonzalez

Meticulous Financial Analyst who undertakes complex assignments, meets tight deadlines, and delivers superior performance — with practical knowledge in corporate finance and financial markets.

Always leading with an ethical and responsible approach, with a vision oriented to results. I thrive in fast-paced environments, tracing clear objectives for my team and myself while fostering healthy competition and collaboration.

Native Spanish and English speaker, fluent in French, and intermediate Chinese. Passionate about continuously expanding my skillset and finding new opportunities to grow.

Financial AnalysisFinancial Modeling Portfolio ManagementData Science & AI Corporate FinanceStrategic Planning Risk AnalysisBusiness Intelligence Market AnalysisClient Relations
4
Languages: Spanish, English, French, Chinese
78.6%
Benchmark outperformance rate in portfolio management
3
Core domains: Finance, Data Science & Consulting

"With a good perspective on history, we can have a better understanding of the past and present, and thus a clear vision of the future."

— Carlos Slim Helú
Global Reach

Experience Across Multiple Jurisdictions

9Cities & Regions
5Countries
3Continents

I've worked across diverse financial markets, legal frameworks, and business cultures — from offshore jurisdictions to emerging markets — bringing a truly international perspective to every engagement.

🇲🇽 Mexico City 🇺🇸 New York City 🇺🇸 San Diego 🇺🇸 Los Angeles 🇬🇧 London 🇮🇲 Isle of Man 🇵🇦 Panama City 🇮🇳 Mumbai 🇪🇸 Madrid, Spain
What I Offer

Services

💼

For Clients & Investors

End-to-end financial advisory and analytical services to grow and protect your capital.

🤖

For Employers & Teams

Technical expertise to transform raw data into strategic advantage.

Selected Work

Projects & Case Studies

💼 Professional
🖥️

Portfolio Management Software

Built software to automate visualization and computation for optimal portfolio selection using graph theory, clustering, and log regression.

⏱ Saved avg. 4 hrs/client/advisor
PythonRBloombergFama-FrenchGraph Theory

The Challenge

Advisors were spending 4+ hours per client manually computing optimal portfolio allocations, creating bottlenecks and limiting scalability for the firm.

Data & Approach

Used Bloomberg Terminal data for historic returns and macro/micro indicators. Applied the Fama-French 5-factor model and Efficient Portfolio Theory, then automated selection using graph theory and log regression in R, later refactored in Python for performance.

Insight & Impact

The software reduced advisor time per client from 4+ hours to under 30 minutes in early-stage deployment. By eliminating manual redundancies, the firm could serve significantly more clients with the same headcount, directly improving revenue per advisor.

📊

Portfolio Management & Client Advisory

Built bespoke investment portfolios for individual clients using Fama-French 3 & 5-factor models and behavioral finance principles.

📈 Outperformed benchmark 78.6% of the time
ExcelPower BIFama-FrenchPsychology

The Challenge

Clients needed personalized investment strategies that balanced their risk tolerance, liquidity needs, and return expectations — while remaining competitive against market benchmarks.

Data & Approach

Leveraged in-house proprietary financial software alongside Excel and Power BI. Applied cohort modeling, the Fama-French 3 and 5-factor models, and behavioral psychology insights to tailor each portfolio to the client's unique risk-return profile.

Insight & Impact

Achieved benchmark outperformance in 78.6% of cases. By combining quantitative factor models with behavioral finance insights, clients not only received better returns but also stayed committed to their strategies during volatile periods — reducing costly emotional decisions.

🏗️

Real Estate Development Analysis

Built a 91% accurate Log-Linear Regression model to determine the optimal apartment mix maximizing profit in a development complex.

💰 +27.65% profit vs. original estimate
MySQLR StudioEDALogistic Regression

The Challenge

A real estate developer needed to determine the ideal ratio of luxury vs. basic apartments in a new complex to maximize overall project profitability, without clear data on what local demand would support.

Data & Approach

Scraped and analyzed sales data from similar complexes in the area using MySQL, then applied Exploratory Data Analysis and a Log-Linear Regression model in R Studio to quantify price drivers and predict demand for each unit type.

Insight & Impact

The 91% accurate model identified the optimal apartment mix and key price-influencing factors. Implementing the data-driven recommendation generated 27.65% more profit than the developer's original estimate, validating the value of applying data science to real estate decisions.

🎓 Academic
🩺

Pima Indians Diabetes Analysis

Exploratory and statistical analysis of diabetes prevalence in the Pima Indian tribe examining genetic and environmental lifestyle factors.

Descriptive StatsEDAData Visualization

The Challenge

Diabetes affects millions globally, yet its root causes remain partially unknown. The study aimed to identify which measurable health factors most strongly predict diabetes onset in a well-defined cohort.

Data & Approach

Dataset comprised female Pima Indian patients aged 21+. Applied descriptive statistics, EDA, and data visualization techniques to analyze distributions and correlations between variables like glucose levels, BMI, insulin, and age.

Insight & Impact

The analysis revealed the strongest predictors of diabetes onset in this cohort, providing a foundation for early screening models. Key findings around glucose concentration and BMI thresholds can inform preventive healthcare protocols.

🏨

Hotel Booking Cancellation Prediction

89% accurate classification model using Decision Trees and SVM to identify customers likely to cancel reservations.

📈 Potential 30% revenue improvement
PythonSVMDecision TreesRandom Forest

The Challenge

Hotel revenue is severely impacted by late cancellations that leave rooms unfilled. The goal was to predict which bookings would cancel, enabling proactive interventions like overbooking or targeted retention offers.

Data & Approach

Used a hotel bookings dataset with 32 features. After EDA and feature engineering, compared Logistic Regression, SVM, Decision Trees, and Random Forest models in Python. Decision Trees and SVM yielded the highest accuracy at 89%.

Insight & Impact

The model identifies at-risk bookings early, enabling hotels to implement targeted retention strategies or adjust overbooking thresholds. Projected revenue improvement of 30% demonstrates the significant business value of applying ML to hospitality operations.

📢

Working Through Ad Block

Assessed the optimal strategy for a company losing ad revenue to ad-blocking users using A/B testing, sentiment analysis, and HR analytics.

💡 $42.8K net productivity increase
A/B TestingSentiment AnalysisMROCHR Analytics

The Challenge

A media company was losing significant ad revenue as users increasingly adopted ad-blocking software. They needed a data-driven strategy to recover lost revenue without alienating their user base.

Data & Approach

Conducted user surveys and MROC (Market Research Online Community) sessions, ran A/B tests on ad delivery strategies, applied sentiment analysis to user feedback, and used HR analytics to evaluate team productivity impacts of different hiring and training paths.

Insight & Impact

Recommended hiring a dedicated ad-block strategy manager combined with a structured team training program. The data-backed proposal projected a $42,800 net productivity increase, providing a clear ROI case for leadership to act on.

🎬

Netflix Movie Recommendation System

Built a collaborative filtering recommendation engine using matrix factorization to suggest movies based on user rating history.

Collaborative FilteringMatrix FactorizationRecommender Systems

The Challenge

Streaming platforms face a discovery problem — with thousands of titles, how do you surface the right movie for the right user? The goal was to build a system that personalizes recommendations from historical interaction data alone.

Data & Approach

Used a movie ratings dataset with historical user-item interactions. Applied collaborative filtering via matrix factorization to learn latent user preferences and item features, predicting ratings for unseen movies and ranking recommendations.

Insight & Impact

The system successfully generated personalized movie recommendations based purely on rating patterns, demonstrating that matrix factorization can capture non-obvious user taste clusters — a core technique powering real-world systems at Netflix and Spotify.

☕

Starbucks Nearest Location Finder

Geolocation model identifying the nearest Starbucks via GPS coordinates or zip code using MongoDB classification and filtering.

MongoDBClassificationGeolocation

The Challenge

Build a user-friendly location tool demonstrating geospatial querying capabilities — finding the nearest point of interest from any given location input, whether GPS coordinates or a postal code.

Data & Approach

Loaded Starbucks location data into MongoDB, leveraging its native geospatial indexing. Implemented classification and filtering logic to accept either GPS coordinates or zip code inputs and return the nearest matching location.

Insight & Impact

Demonstrated practical application of NoSQL geospatial queries — skills directly transferable to retail site analysis, supply chain optimization, and customer proximity modeling in business contexts.

Interactive Tool

Retirement Savings Calculator

Estimate Your Retirement Nest Egg

Adjust the inputs to see how your savings can grow over time with compound interest.

$1,227,405
Estimated portfolio value at retirement

For illustrative purposes only. Past performance does not guarantee future results.

Interactive Tool

Portfolio Optimizer (Markowitz)

Adjust expected returns, volatilities, and correlations for three assets. Then choose to maximize Sharpe ratio or minimize variance.

Asset A — Equity
Asset B — Growth
Asset C — Bond
—
Asset A
—
Asset B
—
Asset C
—
Exp. Return
—
Volatility
—
Sharpe Ratio

Optimization based on historical assumptions. Not financial advice.

Client Stories

Testimonials

"

Cesar's portfolio optimization model saved our team countless hours every week. The automation he built is now a core part of how we serve clients.

A
Advisory Client
Wealth Management Firm
"

The real estate analysis Cesar provided was eye-opening. His model identified opportunities we had completely missed, and the outcome exceeded our projections significantly.

R
Real Estate Developer
Property Development Group
✨

More client stories coming soon — contact me to be featured here.

Contact

Get In Touch

Open to consulting engagements, full-time roles, and interesting collaborations.
Feel free to reach out.

✉️ [email protected] 💼 LinkedIn
📅 Schedule a Call