Data operations and analytics professional based in New York. I build the infrastructure, reporting, and analytics that turn raw business data into clear decisions.
Daily news pipeline pulling AI and tech headlines, scoring sentiment with TextBlob, and extracting named entities with spaCy. Data stored in Supabase PostgreSQL and visualized in a live interactive dashboard.
Financial securities validation system built on PostgreSQL. SQL quality rules across five data dimensions, Pandera schema enforcement, SLA breach detection, vendor scorecard, and a live Streamlit dashboard.
Star schema data warehouse for a simulated hedge fund tracking funds, securities, positions, and transactions in PostgreSQL. Five analytical queries proving the schema supports real business questions.
dbt project modeling raw food delivery data through staging and mart layers with automated tests and lineage documentation. Built on DuckDB with full dependency graph visualization.
Python pipeline using Microsoft Presidio and spaCy to detect, mask, and report on personally identifiable information across structured and free-text fields. Produces a structured compliance report.
Simulation of the human data labeling workflow behind AI model training. Flags invalid labels and low confidence scores across synthetic annotator data with a per-annotator performance scorecard.
Streamlit dashboard visualizing Pandera schema validation results with summary metrics, failure counts, a drill-down table of bad records, and column-level filtering.
Automated schema validation pipeline using Pandera enforcing data type, nullability, and value rules across 12 columns. Deliberately injects bad records to verify failure detection works correctly.
SQL validation pipeline running 10 automated data quality checks against the Titanic dataset using SQLite and Python. Results compiled into a single pass/fail table across five quality dimensions.
Foundational data quality audit of the Titanic dataset using Python and Pandas covering completeness, uniqueness, validity, consistency, and shape across 891 records and 12 columns.
Power BI dashboard analyzing sales and profitability across countries and products for a fictitious company. Compares year-to-date vs. prior year to surface underperforming regions.
Analyzed 1,000+ customer reviews from 2016–2023 in Tableau. Surfaced patterns in service quality, aircraft performance, and geographic satisfaction distribution.
Centralized cost data across 34 NYC residential buildings replacing manual spreadsheets with automated reporting to support budgeting and decision-making.
Personal website for J. Alexander Martin, co-founder of FUBU. Built during my time at Vanguarde, the site covers his work across fashion, tech, media, and business — with sections for consulting, ventures, press, and speaking.
A personal knowledge system with audio memo capture and an interactive knowledge map. Take notes, record voice memos, and watch your ideas connect into a visual graph you can actually navigate.
A minimalist daily task manager built with PyQt5. Designed around focus and intentionality — set one Most Important Task (MIT) for the day, maintain a flexible might-do list, and cut through the noise. Dark-mode interface with automatic date tracking and local storage.
An interactive map cataloging murals across New York City for Thrive Collective. Lets users explore the city's public art scene geographically — find murals by neighborhood, artist, or location. Demo available on request.
A breakdown of ISIN, CUSIP, and SEDOL identifiers — why they matter, where they break, and how data quality failures cost firms millions.
How analysts can implement Great Expectations and Pandera to catch data issues before they become business problems — no dedicated engineering team required.
Behind every aligned LLM is an operations layer most people never see. This maps the systems, roles, and quality mechanisms that make human feedback work at scale.
Data operations and analytics professional based in New York. I build the infrastructure, reporting, and analytics that growing businesses need — without the complexity or the wait.