Built a SQL-based ETL pipeline that pulls data from multiple sources into a single client-level reporting system, making it much faster to monitor and analyze performance metrics as they happen.
Added automated email alerts driven by SQL logic, so stakeholders get flagged the moment a metric crosses a threshold instead of finding out after the fact.
Audited the underlying datasets and fixed formula errors and reporting inconsistencies along the way, so the numbers everyone was making decisions on could actually be trusted.
SQLETLAutomation
Shelfie
A social reading app that lets you search millions of books in real time by pulling live results from the Google Books API straight into a Supabase-backed catalog.
Designed the database schema to support shelves, friendships, and recommendations, with personalized feeds built through server-side data aggregation instead of a generic activity feed.
Built as a proper multi-user app on Next.js App Router with Supabase SSR authentication, so every user's shelf and account data stays private to them.
Trained a Python and Scikit-learn model to flag AI-generated text, spending most of the effort on preprocessing and feature extraction to get prediction accuracy up.
Wired the model into a Flask backend for inference, using Pandas, NumPy, and Joblib to keep text processing fast enough for a live app.
Wrapped it all in a simple web interface so non-technical users get a clear, factor-by-factor breakdown of why a piece of text was flagged, not just a yes/no verdict.
A full-stack scheduling app built in React that pairs real-time task management with location-based services to help students and newcomers find their footing around the city.
Synced live with users' Google Calendars via OAuth, so events created or updated in the app show up on their calendar and vice versa.
Took 1st place for “Best Community Business Solution” at The Golden Hack 6.0 for connecting people to local essential services in a way that actually made daily life easier.
An automated grading tool built in Excel with VBA, using SQL queries under the hood to pull, transform, and analyze student data without manual spreadsheet wrangling.
Linked Excel to Word to auto-generate statistical summaries, charts, and KPIs directly from the grading data, cutting a repetitive reporting task down to one click.
Wrote setup and maintenance docs so instructors could deploy and run the tool themselves, without needing a developer around every time something changed.