SEO Operations System
A seven-tab tracker, three Python automations, and a documented SOP library put a multi-client SEO portfolio in one place. Recurring checks, onboarding, and reporting stopped living in scattered spreadsheets and became a system people could follow.
What I Built
- Seven-tab SEO master tracker
- Three Python automation scripts
- Documented SOP library
- Repeatable client onboarding structure
- Weekly reporting workflow
Schema Markup Audit & Generator
A public tool that audits a live URL, explains what structured data is present or missing, and generates ready-to-paste JSON-LD. It uses Astro for the interface, a Cloudflare Worker for the API, and the Claude API for recommendation logic without pretending schema guarantees rankings.
What I Built
- Live-URL schema audit
- Ready-to-paste JSON-LD output
- Astro front end and Cloudflare Worker API
- CMS-specific implementation instructions
- Anonymous, email, and paid usage tiers
Scalable Dealer Locator
A North American dealer locator built to grow from nine Canadian provinces to more than 50 regions. The custom WordPress data model replaced a broken static page without adding a third-party map platform.
What I Built
- Custom dealer post type
- Hierarchical dealer taxonomy
- CMB2 fields for dealer metadata
- Custom-theme finder interface
- No SaaS map dependency
Ranking-Drop Forensics Workflow
An 82% collapse in daily impressions looked like one SEO problem. Search Console, backlink, crawl, and host evidence separated three overlapping causes that needed three different fixes.
Diagnostic Evidence
- Search Console timeline analysis
- Backlink master-list audit
- WordPress database evidence
- Host and crawl diagnostics
- Cause-specific remediation plan
Automated Analytics Reporting
A 637-line Python pipeline connects to the GA4 and Google Search Console APIs, pulls the data, and writes a dark-themed HTML report. Windows Task Scheduler handles weekly delivery without a manual export.
What I Built
- GA4 Data API and GSC API integration
- Data-based marketing insights
- Dark-themed HTML report generation
- Topic cluster detection from search queries
- Windows Task Scheduler delivery
Multi-Agent Content Pipeline
A six-stage Python and Streamlit application moves a brief through research, outlining, drafting, voice, and QA. Human checkpoints stay in the workflow, and prompt versioning keeps the output reproducible.
What I Built
- Six sequential workflow stages
- Python and Streamlit application
- Brand-voice preflight
- Human review checkpoints
- Human-only and AI-assisted paths
This is what the technical work looks like after it ships.
The case studies lay out the problem, the build, and the result so you can see what each system actually did.
Need a system that does the job?
I build technical SEO workflows, reporting pipelines, custom tools, and the web infrastructure around them. Tell me what keeps breaking or taking too long.