How I Operate · Platform Independence

Platform experts break when the platform changes

Process vs. Procedure

I've watched engineers become useless the moment the toolset changed. Not because they weren't skilled — because they had fused the method to the tool. I keep them surgically separate. The process never changes. The procedure just speaks whatever language the stack speaks.

// the part that never changes

Process

The consistent method I apply regardless of environment, client, or toolset. Triage → Validate → Remediate → Verify → Document.

This is what I bring to every engagement. It doesn't care what RMM you're running or what EDR is deployed. The spine is always the same.

// the part that adapts

Procedure

The tool-specific execution of each process step. Which dashboards, which log queries, which API calls, which console paths.

This changes with every client. ConnectWise looks different from NinjaOne. SentinelOne looks different from Huntress. The procedure speaks the local dialect.

How the separation plays out · step by step

Step 01

Triage

// process (constant)

Confirm the symptom is real, establish scope, prevent premature action

// procedure (adapts)

Which dashboards, health pages, and basic connectivity tests exist in this environment

Step 02

Validate

// process (constant)

Corroborate findings with independent sources, build a defensible timeline

// procedure (adapts)

Which logs, queries, and telemetry sources exist for this specific client and toolset

Step 03

Remediate

// process (constant)

Smallest viable change, pre-check, rollback defined, staged where possible

// procedure (adapts)

The exact policy change, script, exclusion, or configuration adjustment in this platform

Step 04

Verify

// process (constant)

Prove the fix worked, prove no regression, validate full cycle where needed

// procedure (adapts)

How success is proven inside this platform and independently confirmed outside it

Step 05

Document

// process (constant)

Create closure that a peer can follow cold — no tribal knowledge required

// procedure (adapts)

Format and location: ticket notes, KB article, assessment writeup, evidence pack

Keeping these separate prevents two failure modes I've seen destroy teams: tool lock-in — where engineers can't function outside their comfort stack — and tribal knowledge dependency — where the method only exists in someone's head and walks out the door with them. When process and procedure are separate, any engineer can own the process and learn whatever procedure the environment requires.