Data Scientist
in the Making

Hello! I'm Sharon Joanna Azariah, currently pursuing my M.S. in Data Science at George Washington University. I'm originally from India, and I earned my B.E. in Computer Science Engineering.

I enjoy building end-to-end data solutions—from cleaning and analyzing messy data to modeling and communicating insights that support real decisions, especially for impact and the public good.

Outside of data science, I enjoy journaling, watching movies, and baking.

Read more →
View ResumeExplore ProjectsContact
Sharon Joanna Azariah

What I Build

End-to-end data solutions across the entire pipeline

1

Extract

PythonSQLAPIs
2

Clean

PandasNumPyOpenRefine
3

Model

Scikit-learnTensorFlowXGBoost
4

Visualize

TableauPower BIMatplotlib
5

Deploy

DockerFastAPIMLflow

Demo: Healthcare Claims Denials Analytics

This dashboard is a UI demo that shows how I approach analytics for healthcare revenue-cycle operations. The goal is to understand denial trends, identify top denial reasons, and prioritize operational fixes that reduce rework and speed reimbursement.

Problem: Denials create revenue leakage and increase manual follow-up.
Analysis goal: track denial rate over time and pinpoint highest-impact drivers.
Outcome: stakeholder-ready charts with clear actions.

Data: sample dataset (UI demo) — created for visualization only.

How I approach analytics

1Define the question
2Pull reliable data
3Validate
4Visualize
5Explain the 'so what'

I focus on clarity, data quality, and actionable recommendations for stakeholders.

Sample visualization (UI demo)

Denied Claims Rate (%) — Monthly

Insight: Denial rate rises in Q3. Next step is to drill into denial reasons and payer mix to pinpoint operational fixes.

SQL → Insight

Sample workflow (UI demo) — query patterns and how results get visualized.

Demonstrates analyst/analytics engineer capability: SQL + narrative + visualization.

Business Question

How is the denial rate changing month over month?

Sample SQL (UI demo)

SELECT 
  DATE_TRUNC('month', service_date) AS month,
  COUNT(*) AS total_claims,
  SUM(CASE WHEN denial_code IS NOT NULL THEN 1 ELSE 0 END) AS denied_claims,
  ROUND(SUM(CASE WHEN denial_code IS NOT NULL THEN 1 ELSE 0 END) * 100.0 / COUNT(*), 2) AS denial_rate
FROM claims
WHERE service_date >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY DATE_TRUNC('month', service_date)
ORDER BY month;
Sample output (UI Demo)

Insight: Denial rate rises in Q3. Next step is to drill into denial reasons and payer mix to pinpoint operational fixes.

Why it matters: This helps prioritize operational fixes that reduce rework and speed reimbursement.

What you get when you work with me

Clean, validated datasets

I focus on data quality so decisions are built on reliable foundations.

Dashboards stakeholders trust

Clear visualizations that tell the story behind the numbers.

Clear communication

I explain technical findings in terms stakeholders can act on.