Skip to content

EMBER FIELD

  • › initialising flow field…
  • › compiling curl shader
  • › allocating particle buffers
  • › warming transform feedback
  • › igniting the furnace
SL

[ INDEPENDENT SPECIALIST — INFRASTRUCTURE · PLATFORMS · AUTOMATION ]

StefanLarsen

I build the systems that keep other systems running.

I automate the operational work between them — improving infrastructure, platforms and automation so complex IT environments become easier to operate, scale and trust.

Let’s talk about your problem
See how I work

// SOUND FAMILIAR?

Your IT works. But it shouldn’t be this hard.

  • Too many processes still done by hand
  • Systems that don't talk to each other
  • Infrastructure that has become difficult to maintain
  • Critical processes that depend on tribal knowledge
  • Identity and access workflows that are still manual
  • Monitoring that says something is wrong — but not what to do
  • A platform that grew organically and now needs structure
  • A project that needs a specialist for a few months
Have one of these problems? Let’s talk

// WHAT I HELP WITH

01

Infrastructure Automation & Auto-Remediation

MONITORDETECTGOVERNEXECUTEVERIFYREPORT

The problem
Infrastructure alerts still trigger manual operational work: investigation, a remote session, remediation, documentation, escalation — across monitoring, ITSM and endpoint tooling, by several teams.
What I do
I connect monitoring, ITSM, orchestration and endpoint tooling into governed automation workflows — so an alert triggers the right remediation, verification and documentation without a person working every step.
Outcome
Repetitive operational work moves out of the NOC/L2/L3 queue into an automated workflow — with change governance, diagnostics and an audit trail built in.

As built: Zabbix alert → webhook → ServiceNow change → BigFix → endpoint remediation → verified result → work note

  • Zabbix
  • ServiceNow
  • BigFix
  • PowerShell
  • .NET
Proof: Disk Cleanup Automation
02

Automation & Identity

The problem
Every hire, leaver and access change is a ticket someone works by hand — across several systems, by several teams.
What I do
I automate the identity lifecycle end to end — directory, licensing, mailboxes — and put a self-service front on it.
Outcome
Requests run without an engineer in the loop: faster, consistent, auditable.
  • ServiceNow
  • Active Directory
  • Entra
  • Exchange
  • PowerShell
03

Integration & Systems

The problem
Data lives in systems that don't talk, so people re-key it, export it, and reconcile it by hand every cycle.
What I do
I build the integrations and pipelines between them — APIs, databases, reporting — running unattended on a schedule.
Outcome
The manual hours disappear; reporting reads from live, structured data instead of spreadsheets.
  • REST APIs
  • Node.js
  • MySQL
  • Power BI
04

Infrastructure & Platforms

The problem
The platform grew one urgent fix at a time. Every customer or environment is a snowflake, and changes are risky.
What I do
I extract the patterns into a governed, reusable platform — shared engines, credential handling, change control.
Outcome
New environments become configuration instead of projects, and operations become explainable to an auditor.
  • ServiceNow
  • BigFix
  • Azure Key Vault
  • CMDB

// REAL PROBLEMS I'VE SOLVED

MONITORING-DRIVEN AUTO-REMEDIATIONExample: disk spaceZabbix alertwebhookServiceNow changeBigFixendpoint remediationdiagnosisverified resultwork noteThe alert doesn’t just get fixed: the change is governed, the disk diagnosed, recommendations written back, everything audited. Disk Cleanup Automation is one example of the pattern — open the case SEE AUTOMATION IN ACTIONAccess changes and infrastructure alerts worked by hand mean delays and errors. This is the engine that does it instead — detection to verified result. Drive a real flow yourself IDENTITYHybrid Identity AutomationProblem Every hire, leaver and change was a hand-worked ticket — per tenant, per engineer.What changed One multi-tenant self-service engine across AD, Entra and Exchange.Result Requests run end to end without an engineer; onboarding a customer is config, not code.MONITORINGMonitoring Platform ConsolidationProblem Three monitoring platforms — three infrastructures, alerting schemes and escalation paths.What changed One Zabbix estate across five countries, with automated deployment and auto-registration.Result A single pane of glass; the cost and overhead of two whole platforms retired.COMPLIANCECompliance Audit AutomationProblem Security-baseline audits were assembled by hand on production servers — slow, inconsistent, and itself a risk.What changed A self-service, read-only CIS scan delivered through the automation engine.Result A structured, repeatable result in minutes — trended per server and customer, with no privileged logins.
All 10 systems Have a similar problem? Tell me about it

// HOW I WORK

  1. 01

    Understand

    We start with the problem, the existing environment, and what's currently getting in the way.

  2. 02

    Map

    I identify the dependencies, bottlenecks and manual processes — and the most valuable improvements.

  3. 03

    Build

    I design and implement inside your existing environment, rather than rebuilding everything for its own sake.

  4. 04

    Hand over

    You get something your own team can understand, run and maintain — documented, not dependent on me.

5+years building & operating complex IT systems
10systems taken into production
5countries — international delivery & operations

Built and operated on an estate shared by banks, military and government — environments where “probably fine” isn’t a state you’re allowed to ship.

I work as an independent specialist and deliver remotely — through the same channels your teams already use: tickets, repos, docs and calls.

Have a difficult IT problem?

Tell me what you’re trying to solve. I’ll tell you if I think I can help — and if I can’t, I’ll say so.

Start a conversation