DATA
Open to full-time 2026

Joash / Data Analyst

Data Analyst with hands-on business ownership experience. Applied SQL, Python, and BI tools to real P&L decisions across hospitality, healthcare, finance, and HR — delivering ML pipelines, automated models, and KPI insights that drive operational impact.

View Projects Contact Me
01 — About

I speak
both languages.

I'm a data analyst who bridges the gap between raw data and real decisions. My work lives at the intersection of rigorous statistical analysis and clear, compelling storytelling — because insights that can't be communicated don't drive change.

With an MBA in Finance & HR, and hands-on business ownership in hospitality, I've built ML pipelines in Databricks, automated Excel/VBA models, and KPI dashboards that directly informed P&L decisions across healthcare, finance, and operations.

I believe the best analysis asks a better question, not just finds a better answer. I thrive where ambiguity is the norm and clarity is the goal.

Data & Programming
SQLPythonPandasNumPyScikit-learnR
Analytics & BI
TableauPower BI (DAX)ExcelPower QueryVBA
Big Data & ML
PySparkDatabricksSnowflakeKNIMEETLAPI Integration
Modeling & Statistics
RegressionRandom ForestKNNTime SeriesMonte CarloA/B TestingNLP
02 — Projects

Work that
moves numbers.

Each project starts with a real business problem and ends with a measurable outcome.

01
HealthcareExcel/VBAScenario Modeling

ER Visit Reduction — HCE Case Study

Designed a fully automated Excel/VBA model for multi-scenario cost and savings analysis, eliminating manual recalculation and improving evaluation efficiency. Modeled 3+ intervention scenarios to identify the highest-value cost-reduction strategies for decision-makers.

Intervention Scenarios
4
cost-reduction paths modeled
ExcelVBAScenario Analysis
02
HealthcarePySparkML Pipeline

Gender Bias in Medical Outcomes

Built scalable PySpark ML pipelines in Databricks to process large clinical datasets, evaluate classification model performance, and surface statistically significant gender-based bias patterns with corrective recommendations.

Pipeline
Scalable
PySpark ML in Databricks
PySparkDatabricksClassification
03
FinancePythonML Comparison

Data-Driven Investment Model

Built and compared 3 ML models (Random Forest, KNN, Linear Regression) with feature engineering across multi-variable financial datasets to optimize forecast accuracy. Random Forest outperformed baseline assumptions after cross-validation, identifying key performance drivers across asset classes.

Best Model
8.7%
lower RMSE after cross-validation
PythonRandom ForestKNNScikit-learn
04
HealthcareKNIMEVisual Analytics

AI Impact on Radiologist Efficiency

Built KNIME workflows and Tableau dashboards to analyze AI-assisted diagnostic tool impact on radiologist speed and accuracy across large-scale radiographic datasets.

Analysis Surface
11K+
AI-assisted imaging records analyzed
KNIMETableauWorkflow Automation
05
EducationRNLP

Emotional Intelligence as a Predictor of Academic Success

Transformed unstructured student sentiment data into quantifiable behavioral indicators using R and NLP; regression analysis demonstrated a significant EI-to-performance correlation.

Finding
0.68
correlation coefficient observed
RNLPRegression
03 — Data Insights

Signals in
the noise.

ER Visits — Scenario Cost Savings
Combined intervention strategy projected 31% reduction in avoidable ER utilization
Base
Int. A
Int. B
Int. C
A+B
B+C
All
▲ Automated Excel/VBA model evaluated 3+ intervention mixes for HCE decision-makers.
Investment Model — Predicted vs Actual
Random Forest achieved the lowest prediction error after cross-validation
Model trained → Jan Jul Dec
Random Forest
KNN / Linear
Confidence
Radiologist Workflow — AI Impact
AI-assisted workflows improved turnaround efficiency across diagnostic reviews
11.2K reads
34%
Complex cases
29%
Routine — AI-assisted
23%
Review / QA
14%
Admin
Gender Bias — Classification Gap
Bias evaluation identified measurable recall disparities across cohorts
Accuracy
89%
Precision M
84%
Precision F
76%
Recall gap ⚠
-9pt
Post-fix
91%
04 — Contact

Let's build
something.

Open to full-time Data Analyst / BI roles. If you're turning ambiguous business questions into decisions, let's talk.

jnk.analyst@gmail.com