Hi, I'm Daniel.

A data nerd with a strong passion for transforming complex datasets into clear visual stories and reports that allow teams to make informed decisions. I focus on creating results that actually enhance understanding and drive operational efficiency through data.

My work spans several industries, such as:

• Retail
• Real Estate
• Marketing
• Finance
• Research
• Geospatial Analysis

This breadth has given me a versatile, cross-disciplinary perspective on solving data problems, along with the adaptability to get up to speed quickly in unfamiliar domains.

Projects

Airline Route Network Analysis

Open in browser

Graph-traversal analysis of the global airline route network, covering 3,300+ airports and 67K+ routes, built using DuckDB and recursive CTEs, with DBeaver for schema browsing and validation. Finds shortest hub-to-hub paths between any two airports, ranks airports by one-hop connectivity to surface the busiest hubs, and simulates removing a major hub to score how much reachability the network loses. Found Frankfurt, Paris CDG, and Amsterdam as the most connected hubs, with most airport pairs reachable in two or fewer stops.

DBeaver Docsify DuckDB Recursive CTEs SQL

BVC Analytical & Predictive Model

Open in Google Data Studio

Made a Google Apps Script scraper that feeds Google Sheets with double SHA-256 deduplication, powering a Google Data Studio dashboard with live market and technical-analysis views for the Venezuelan Stock Exchange (Bolsa de Valores de Caracas). A KNIME workflow then trains Random Forest and Gradient Boosting models to forecast closing prices, reaching a MAPE as low as 3.6%.

Google Apps Script Google Data Studio Google Sheets KNIME Machine Learning Web Scraping

Chicago Crime Hot Spots Analysis

Open in browser

A spatial hot-spot detection pipeline on 500K+ Chicago crime records built using PostgreSQL and PostGIS, binning incidents into a SQL-generated 500-meter grid and scoring concentration with window-function z-scores against citywide monthly averages, alongside a 90-day trend comparison to flag emerging hot spots. Found dense, statistically significant clusters along a corridor through the South Side and a secondary concentration near downtown, with several cells showing sharp short-term increases.

Geospatial Analysis PostGIS PostgreSQL SQL Window Functions

‹ ›

Crude Oil Price Shock Analysis

Open in Loom

An interactive Power BI dashboard tracing WTI and Brent crude prices across four decades (1986-2026), pairing forward-filled FRED/EIA price series with a timeline of 27 geopolitical and market shocks. DAX measures drive a rolling 30-day volatility model with percentile-based thresholds, a dynamic event-window filter, and an events table spanning everything from the 1986 OPEC price war to the 2026 Iran conflict.

DAX Data Modeling Data Visualization Power BI Power Query

Data Quality Monitoring Pipeline

Open in KNIME

A KNIME pipeline that ingests 25 monthly batches of transactional data, profiles each against a fixed baseline, and flags it OK, WARNING, or CRITICAL using schema checks, drift metrics, duplicate-row detection, and an H2O Isolation Forest for anomalies. Thresholds were calibrated against 22 clean batches rather than guessed, and the pipeline caught all three corrupted batches plus one real, unplanned duplicate-row spike.

Anomaly Detection Data Quality ETL KNIME Machine Learning

NYC Airbnb Market Analysis

Open in Tableau

Wrote a Python script using geopy's Nominatim client to reverse-geocode nearly 49,000 listings from lat/lon into clean addresses and ZIP codes, respecting the rate limit so the run can pause and resume safely. The enriched dataset powers a Tableau dashboard exploring neighborhood pricing, occupancy rate (a calculated field), and geographic trends with interactive filters across all five boroughs.

Data Caching Excel Market Analysis Python Reverse Geocoding Tableau

NYC Congestion Pricing Analysis

Open in browser

Built a before/after SQL analysis of NYC's January 2025 congestion pricing rollout, comparing 22M+ taxi trips across matched six-month windows using DuckDB, with a recursive CTE for gap-free daily trends and window functions for month-over-month and zone-ranking shifts. Found an 11% overall rise in trip volume alongside a sharp divergence between Manhattan and the outer boroughs. Published the results as a lightweight SQLite extract with a live, browser-based SQL query tool.

DuckDB SQL SQLite Urban Mobility Window Functions

S&P 500 Live Data Dashboard

Open in Google Data Studio

Sources live prices and fundamentals for all S&P 500 current stocks straight into Google Sheets via Google Finance, with custom LAMBDA and MAP formulas handling per-ticker lookups and rolling calculations without using Apps Script. The sheet refreshes automatically and feeds a dashboard with interactive filters tracking relevant KPIs, sector performance, YoY price and market cap trends.

Automation Financial Analysis Google Data Studio Google Finance Google Sheets

The Fallen of the Napoleonic Wars

Open in browser

An interactive isotype chart visualizing military and civilian deaths across the Napoleonic Wars (1792-1815), rendered entirely in vanilla JavaScript and Canvas with no charting libraries. Users can toggle between a yearly view and a by-nation breakdown, filter the timeline with a dual-handle year slider, and sort by casualties. Hovering any symbol reveals a tooltip with exact casualty figures and sourced battle references.

Canvas API Data Visualization Historical Data Interactive Design JavaScript

Skills

Languages