Four studies spanning enterprise platforms, delivery machinery, workflow modernization, and an independent product. Confidential work is presented as redacted studies — real problems and decisions, abstracted visuals, nothing sensitive.
01 / Enterprise internal platform
A multi-tool internal DBA platform that consolidates license auditing, environment metadata, platform monitoring, patch pipelines, and per-client operational views — work that was previously scattered across databases, scripts, spreadsheets, legacy desktop tools, and tribal knowledge.
Redacted study — confidential internal work
Read the study02 / Independent software product
An end-to-end World of Warcraft combat-analytics platform: an in-game Lua addon captures encounter data, a Node/PostgreSQL backend ingests, validates and scores it, and a Next.js site surfaces rankings, player profiles, and a premium tier.
Read the study03 / Enterprise workflow modernization
A client testing portal on Azure DevOps: Excel-based UAT moved into Test Plans, external client users onboarded as Entra ID guests into a repo-less project, and an event-driven pipeline — service hooks → Azure Function → Service Bus → on-prem receiver — that keeps client-facing bugs, internal work items, test cases, and legacy CM cases correlated by ID. A systems design problem wearing a configuration exercise's clothes.
Redacted study — confidential internal work
Read the study04 / Enterprise engineering & technical operations
Making complex Oracle database delivery safer, more visible, and easier to operate — across 11g, 12c, and 19c environments, dozens of schemas, a 20-year-old configuration-management stack (Scopus / PVCS), and a change process where a single careless script can take a client down.
Redacted study — confidential internal work
Read the studyConcepts, prototypes, and research — explicitly not shipped client work.
Exploration of a Quebec-focused contractor marketplace with an integrated CRM — matching workflows, licensing data, and quote management. Product and data-model research stage.
Working with Régie du bâtiment du Québec open data: license records, ingestion, normalization, and what a useful public interface over it could look like.
Small-scale experiments in endpoint polling, health rollups, and low-noise alert design — informed by building the store-status monitoring dashboard at work.
Ongoing analysis of combat-log structures and scoring models feeding future Cadence development — encounter normalization, outlier detection, and fairness in ranking.
Cross-platform React Native app for beer discovery — ratings, barcode scanning, event navigation — backed by a Python pipeline that collects and geocodes Quebec brewery data.