Projects · 16 systems

Systems that operate themselves.

Data and AI systems I designed and built. Filter by stack, pattern, or domain. Every entry links to a case study with an architecture diagram and says plainly whether it is in production, a pilot, or a prototype.

Async ETL Platform
Async multi-pipeline ETL into Postgres: about 14 scheduled pipelines across 25 GitHub Actions workflows
pythonasyncetlpostgresqlmaterialized-viewsincremental-syncautomationsupabasededupdata-quality
● In production
Workforce Report-Compliance Platform
Internal platform for daily-report approvals, reminders, management PDFs and extracts, with write-back to the vendor API as each approver
nextjsreacttypescriptsupabasepostgresqlsecurityautomationpdfdashboards
● In production
Quote Automation
Internal quoting tool that took accounting from 20-30 quotes a day to over 100: review queue, batch PDF-to-Drive and Gmail-draft emailing
nextjsreacttypescriptsupabaseautomationpdf
● In production
Pipeline Guardian
Email-conversational ETL remediation agent with human-in-the-loop approvals
pythonclaude-apiagenttool-useautomationsupabasepostgresqlsecurityai-safetyemail-parsing
● In production
DARA: Data Analyst Reporting Agent
Chat-first NL→SQL analytics with Postgres RLS and a defined-metric library (prototype)
typescriptnextjssupabaseclaude-apinl-sqlagenttool-usepostgresqlrlsai-safety
Prototype, in development
Event-Driven Sync on Google Cloud
Change events over SSE trigger debounced incremental walks and guarded idempotent upserts, with an hourly reconcile as the safety net
pythonpostgresqlincremental-syncetlautomationsecuritydata-quality
● One job in production · listener in shadow
Report Automation
Scheduled report suite: finance, compliance and open-items reports via Gmail, Drive and Google Chat
pythonautomationetlpostgresqlsupabasepdfdashboardsvisualization
● In production
Cross-Source Date Validator
Daily reconciliation of task completion dates in the system of record against dates reported by email
pythondata-qualityautomationpostgresqlsecuritydedupincremental-syncasync
● In production
Gmail Document Parser
HTML email → JSONB with dynamic field discovery
pythonemail-parsingpostgresqlautomationdedupsupabase
● In production
Report Run Ledger and Scheduler
Pilot: one place that records every report run, dispatches scheduled ones, and flags missed or stuck runs
nextjstypescriptsupabasepostgresqlautomationsecuritydashboards
Pilot (test phase)
RFDS Extractor
Gmail PDF scraper with a deterministic parser + optional Claude fallback
pythonemail-parsingpdfclaude-apiautomation
● In production
PDF Attachment Extractor
Org-wide parallel PDF requirement downloader over the vendor's project-management API
pythonpdfautomationetlincremental-sync
● In production
GC Asset Lake
Lakehouse-lite ETL: incremental API sync with hybrid hot (Postgres) / cold (Parquet + DuckDB) storage
pythonetlpostgresqlincremental-syncautomationasync
Prototype
Daily Claude Digest
Personal daily briefing: fail-soft fetchers, one Claude composition call, sent over Gmail, scheduled via GitHub Actions
pythonclaude-apiautomationprompt-engineering
Personal tool
Ops Portal
Internal ops portal UI: unified data-health, dashboards & embedded analyst (foundation phase)
nextjsreacttypescriptsupabasesecuritydashboards
Prototype
Agent Town
Live pixel-art town that visualizes a monorepo's activity in real time
typescriptvisualizationwebsocketsnodejsautomation
Prototype