Self-Hosting for Families: My Unraid Journey
Why I run my own home server with Unraid, and how self-hosting changes your relationship with technology and data ownership.
When I transitioned from Process Engineer to MEGA Regional Project Manager at Mars in 2019, the scope of my work expanded from a single facility to dozens of sites across North America and Latin America. The key to managing that scale wasn’t working harder. It was building better systems for visibility.
Managing MRO (Maintenance, Repair, and Operations) supply across multiple manufacturing sites creates a visibility problem. Each site has its own:
Without systematic visibility, you’re flying blind. Issues hide until they become crises.
My approach was to build transparency through dashboards:
Vendor Compliance Heatmaps
Power BI Efficiency Dashboards
Ariba Reporting Tools
This data-driven approach delivered measurable outcomes:
| Metric | Before | After | Impact |
|---|---|---|---|
| SAE | 74.6% | 84.8% | +14% efficiency |
| Waste Rate | 4.6% | 1.9% | -59% reduction |
| PCC | $2,853/ton | $2,575/ton | -10% cost |
Financial Impact:
Getting to this point required deep technical work:
SAP Classification Mapping: Learning how items are classified in SAP to map across contracted categories. This wasn’t glamorous work, but it was essential.
Inventory Calculator Automation: I developed approximately 50% of a solution to automate the SAP data pull for inventory calculations. Reducing manual work meant more time for analysis.
Data Lake Integration: Working with the data sprint subteam to bring live data into dashboards, replacing constant manual SAP reports.
Network Troubleshooting: When SAP scripting went down across the network, I found the solution and rolled it out to the entire maintenance diagnostics team.
This experience directly informs my perspective on product management:
Build for Visibility: Users don’t just need features; they need to see what’s happening. Dashboards aren’t nice-to-have; they’re core product.
Data Quality is Product Quality: The best analytics are worthless if the underlying data is messy. Invest in data hygiene.
Automate the Boring Parts: Every hour spent on manual data pulls is an hour not spent on analysis and decision-making.
Measure What Matters: SAE, PCC, waste rate: these aren’t arbitrary metrics. They directly tie to business outcomes.
The hardest part wasn’t the technical work. It was getting alignment across functions:
Building dashboards that served multiple audiences while maintaining a single source of truth was the real product management challenge.
Whether it’s MRO supply chain analytics or an AI-powered agentic platform, the principles are the same: build visibility, automate the tedious, measure what matters, and serve multiple stakeholders with coherent systems.
That’s what I’m looking to bring to my next role.