Brian Cervantes Alvarez

Statistician & Data Scientist • R/Shiny • Python • SQL • Educator

I design explainable ML and decision-ready analytics, ship interactive tools teams actually use, and teach the concepts clearly enough that adoption sticks.

Hire Me → Download Resume

Featured Projects • LinkedIn • Skills

How I Deliver Results

Clear problem framing, rigorous validation, and production-minded delivery.

Decision-Focused Analytics

Translate goals into crisp KPIs and hypotheses; design analyses that answer “What should we do next?” not just “What happened?”.

Reliable Data Plumbing

Build auditable R/Python/SQL pipelines with documented assumptions, reproducible runs, and failure-aware checks.

Explainable Models

Train and validate with holdout/CV, emphasize interpretability (GLMs, trees, SHAP-style diagnostics) so stakeholders trust results.

Executive-Ready Visuals

Ship interactive dashboards and reports that are readable at a glance and defensible in detail.

Production & Iteration

Deploy lightweight apps, instrument usage, and iterate based on real-world feedback to sustain impact.

Featured Projects

Each case highlights a problem, my approach, and what it shows about how I work.

Teaching & Adoption

Interactive Teaching with webR

Problem: Static lecture materials slow down comprehension of statistical ideas.

Approach: Built browser-native, no-install lessons using webR and Quarto so students can run code inline.

Demonstrates: Education UX, reproducibility, R/Quarto integration, and clear pedagogy.

See Project →

Analytics App

Shiny App: UK Traffic Accident Trends

Problem: Stakeholders need to explore incident patterns without wrangling data.

Approach: Delivered an intuitive R/Shiny app with filters, dynamic views, and sensible defaults.

Demonstrates: R/Shiny engineering, data modeling, thoughtful IA, and visual best practices.

See Project →

Statistical Modeling

Multivariate Analysis: Wine Quality

Problem: Identify drivers of perceived wine quality across correlated features.

Approach: Applied PCA and supervised models with diagnostics to separate signal from noise.

Demonstrates: Sound inference workflow, feature interpretation, and model validation.

See Project →

LLM + RAG

DataScienceYapper (Client-side LLM + RAG)

Problem: Visitors need fast, private answers about my work without server costs.

Approach: Implemented in-browser LLM with retrieval over site content; no data leaves the client.

Demonstrates: Modern web ML, RAG design, UX, and performance-minded engineering.

Open Chat →

Data Science Toolkit

Core languages, frameworks, and workflows I use on real projects.

R / Shiny

5+ yrs

Production dashboards, tidyverse pipelines, teaching materials, and packages.

Python

4+ yrs

ETL, scikit-learn pipelines, automation, data apps.

Statistical Analysis

4+ yrs

Experimental design, GLMs, uncertainty communication, and diagnostics.

SQL

2+ yrs

Schema design, window functions, performance-minded queries.

Consulting

2+ yrs

Scope, plan, and deliver projects with clear tradeoffs and documented assumptions.

Cloud

2+ yrs

Deploy lightweight services on GCP/AWS; CI-friendly, cost-aware setups.

Ready to discuss your data problem?

Send a note with context and timelines; I’ll reply with a concrete next step.

Email Me LinkedIn

Download Full Resume

Resume

Preview inline or download a copy.

Open in new tab Download PDF

Your browser can’t display the PDF. You can download it here.