From Clinic Sysadmin to IAMOps Practitioner: A Skills Story

Healthcare IT

Picture a systems administrator at a mid-sized clinic network, someone who has spent years keeping logins, shared drives and scheduling tools running. One quiet Tuesday she is asked a question nobody on the team can answer: who exactly is allowed to change the appointment-routing rules, and who gave them that right? She starts digging, and a few weeks later she has a new career interest. This story is invented, but the path it follows is one that many IT and operations people are walking, and it has a name that is gaining ground: IAMOps.

This piece is about skills and careers in that space. It is informational only and does not touch on clinical matters; the setting is simply the administrative and technical side of healthcare organisations, where the question of who may do what in a system matters every day.

What the term actually means

The Ops Atlas run by xOps School carries an entry describing IAMOps in careful language. It calls the word an informal label, not a formal job title, and applies it to authorisation as code, the craft of drafting, testing and improving the policies that control which operations humans, services and bots can perform. Because it is informal, you will rarely see a posting titled “IAMOps Engineer”. You will see its skills spread across cloud security, platform and operations roles.

The first skill: thinking in four questions

The administrator in our story learned to break any access question into four parts. Who is asking? What do they want to do? Which thing do they want to do it to? What is the surrounding situation? In the Atlas’s model, those four pieces make up everything an access decision has to work with. Once she could name them, mysteries that had looked random started to make sense, such as why a colleague could view a report on the office network but not from a hotel.

The second skill: reading policies like a contract

Rules are written in structured formats, and careful reading is a craft. The Atlas lists several traps worth knowing early. A rule that says “allow everything except these items” grants far more than people expect. A condition phrased as “not equal” can match when the value is missing. Some engines quietly skip a rule that breaks, so a safeguard can fail without a sound. Our administrator kept a small notebook of these gotchas, and it became the most valuable thing she owned.

The third skill: measuring instead of guessing

Careers in this area reward people who bring numbers. She learned to compare the permissions a role was given against the ones it actually used. In an example from the Atlas, one role was allowed 180 actions and exercised only 27, a utilisation of 15%. Across a set of 120 roles, a quarter of tidying lifted the typical figure from 22% to 41%. Presenting results like that, in a short chart, changed how her managers listened.

The fourth skill: writing and reviewing in a repository

Instead of clicking through consoles, she began storing rules as files, submitting changes for review and attaching tests. This is the everyday rhythm of modern engineering, and it is covered well in our roadmap for people learning DevOps and applying for remote roles. Comfort with version control and automated checks is probably the most transferable skill on this list.

The fifth skill: communicating trade-offs

Tightening access can slow people down, and good practitioners say so plainly. In the Atlas’s example the time to receive a new permission grew from 36 to 48 hours after cleanup, and the team judged the trade worthwhile. She learned to bring that kind of honest cost to meetings rather than hide it. Colleagues trusted her more because she did not oversell.

Choosing tools without being dazzled

A common early question is which product to learn. The Atlas mentions open-source engines and a graduated cloud-native project for evaluating rules, as well as a newer language created by a cloud provider, but it treats the concepts as more durable than any one tool. Our guide on choosing software that matches how an organisation really works offers a sensible way to compare options without chasing trends.

A learning path she followed

  1. Month one: list every account and automated job, and give each an owner. She discovered eleven with none.
  2. Month two: move one important role into a repository, add a reviewer and a test, and trim one over-generous permission.
  3. Month three: run the first monthly review, set up an exception register and decide how the team would evaluate rules going forward.

This tracks the staged first-quarter plan in the Atlas, though she adapted the details to her workplace.

Making the career case in an interview

Hiring managers tend to ask for evidence, so build a small portfolio. A one-page write-up of an access review you ran, with before-and-after numbers and the exceptions you recorded, says more than a certificate list. Describe what went wrong as well as what went right; the Atlas itself notes that cleanup can break infrequent tasks, and an honest account of a near miss shows judgement. Keep any identifying details of your employer out of it, and describe systems in general terms.

Pitfalls for people changing direction

  • Chasing tools before concepts. A policy language changes faster than the idea of least privilege, which the Atlas traces to a 1975 paper.
  • Skipping the boring inventory. Nothing impresses colleagues like a complete, owned list of identities, and nothing else works without one.
  • Ignoring rare jobs. Quarterly and annual tasks look unused for most of the year. Ask before removing.
  • Working alone. The practice depends on owners across teams agreeing to review. Build relationships early.

How the market looks, honestly

Public evidence suggests demand is real. A cloud security survey from late 2025 found year-old access keys still active for more than half of Google service accounts and nearly 60% of AWS users, and the Atlas notes that a 2025 web-risk ranking put broken access control at number one. None of this guarantees a particular job, but it explains why employers care about people who can reduce that exposure.

Where she ended up

A year later our administrator led an internal review programme and trained colleagues in two other departments. She did not need a new title. She needed a habit of asking who may do what, a notebook of traps and the patience to repeat the loop. If you work anywhere near systems that hold sensitive records, the same path is open to you, one question at a time.