Selected Work

Projects.

A mix of cloud infrastructure, AI experiments, and small tools I've built to learn or solve a personal itch.

One version-controlled source of truth for the personal instructions I give every AI tool, built out to each one so nothing drifts.

Azure API Management placed in front of Azure OpenAI as one controlled front door, with per-client keys, token limits, and caching.

Bicep template that provisions an Azure Storage Account with static website hosting and deploys a site to it in a single command.

Multi-container Docker application that shortens URLs using three networked services: nginx reverse proxy, Flask backend, and PostgreSQL database.

Azure Speech-to-Text pipeline that streams audio transcription and uses an LLM to analyze call sentiment and provide detailed feedback reports.

Automated pipeline for analyzing investor sentiment from earnings calls and financial documents.

Deep learning pipeline built with Claude Code to classify cat breeds using EfficientNetV2.

Machine learning model built in Azure ML Studio to predict flight delays based on historical data.

Lightweight Python tool to access computer files from a phone without cloud services or account signup.