Home › Topics › Monitoring Skills and Certifications for Career Growth › Building a Monitoring Portfolio

Building a Monitoring Portfolio: Evidence of Skill You Can Show


Designers have portfolios. Developers have code they can point to. Administrators have a CV that says "responsible for monitoring". That is the same sentence the previous candidate's CV said. And administrators have an interviewer who has no way to tell the two of you apart. That is a problem you can fix in an evening a month. And the fix is a portfolio: a small set of things that prove you did the work.


A portfolio, for an administrator, is not a website full of screenshots. It is five or six artefacts, each one page. Each shows a real piece of monitoring work with the graph that went with it and a short explanation of what you were thinking. It is small on purpose. An interviewer will look at one artefact, maybe two. The purpose of the portfolio is that when they say "tell me about a time", you can say "here is one", and put a graph on the table.


By the end of this guide you will know which six artefacts belong in a monitoring portfolio. You will know what to leave out and how to sanitise what stays. You will know the one-page write-up shape, where to keep it, and how to use it in an interview without seeming to have rehearsed. And you will have a portfolio index and a sanitisation checklist you can copy.


In this guide

  • What goes in
  • What stays out, and how to sanitise
  • The write-up shape
  • Where to keep it
  • Using it in an interview
  • The portfolio index
  • The sanitisation checklist
  • One graph at Acme
  • The portfolio checklist
  • Where to go next

What goes in


Six kinds of artefact prove monitoring skill. A portfolio with one of each covers every rung of the skill ladder an interviewer is likely to ask about. You do not need all six on day one. Two is a portfolio. Six is a strong one.


  • A monitoring design write-up. It is one page describing a service you designed monitoring for. It covers what it depended on, what "working" meant, which checks you chose and why, which thresholds and why those numbers. This proves you can think about a service rather than only add a ping.
  • A dashboard you built, with an explanation of the audience. A screenshot and a paragraph saying who it was for, what question it answered, and what you left off it and why. The "left off" sentence is the one that impresses, because most dashboards are built by adding. And what dashboards are for explains why the ones that work are built by removing.
  • A runbook. One you wrote, that someone else followed. Include a line saying who followed it and whether they had to ask you anything. Runbook anatomy describes the shape a good one takes; the portfolio needs only one that took it.
  • A post-incident review you led. The timeline, the cause, the actions, and a note about what changed afterwards. This proves you can run the part of an incident that happens after the fix, which is the part most administrators avoid. Blameless reviews covers how to lead one that people will attend twice.
  • A lab exercise log. The table from learning monitoring by lab, with the surprise column filled in. This is the artefact that proves skill when you have not yet had the job. It is the one that most clearly separates a candidate who has done things from one who has read about them.
  • A custom check with its documentation. The script or configuration, what it tests, what it returns, and how it fails safely. The anatomy of a custom check describes what a complete one contains. The portfolio copy should be short enough to read in two minutes and documented well enough that a stranger could run it.

Each artefact proves a different thing, which is the point of having several kinds rather than six dashboards. The design write-up proves thinking. The dashboard proves judgement about audiences. The runbook proves you write for other people. The review proves you can lead. The lab log proves you practise. The custom check proves you can build. An interviewer who has seen all six knows more about you than the CV could say in ten pages.


What stays out, and how to sanitise


Nothing confidential goes in the portfolio, and "confidential" means more than the things marked confidential. It means real device names, real addresses, real usernames, real customer names. It means the name of the organisation if it could be embarrassed by the story, and any graph that shows a weakness an attacker would enjoy. Your employer's monitoring data belongs to your employer. What belongs to you is the fact that you did the work and the way you thought about it. That is what the portfolio shows.


Sanitising is a mechanical job and worth doing mechanically. Replace every hostname with a descriptive fake on the example.com domain, so that fs02.corp.internal becomes fileserver-2.example.com. Replace every address with one from the ranges reserved for documentation, such as 192.0.2.0/24. Those ranges are set aside precisely so that examples never point at anything real. Keep the shape of the names, because the shape is part of the story. A reader needs to know that this was a file server and that was a core switch. They do not need to know what you called them.


A text export of a design or a runbook can be cleaned with one line before you check it by hand. On Windows, in PowerShell:


(Get-Content design.txt) -replace '10\.\d+\.\d+\.\d+', '192.0.2.X' -replace 'corp\.internal', 'example.com' | Set-Content design-clean.txt

On a Linux host, sed -E 's/10\.[0-9]+\.[0-9]+\.[0-9]+/192.0.2.X/g; s/corp\.internal/example.com/g' design.txt > design-clean.txt does the same. Adjust the patterns for your own address ranges and domain. Then read the result yourself, because a one-liner will not catch the customer name in a comment or the username in a screenshot. The one-liner is the first pass. Your eyes are the second, and there is no third.


Fictionalise the organisation when the story could identify it. A post-incident review that says "the clinic's patient records system was unreachable for forty minutes" is a story about a real place. The same review about "a mid-sized organisation's main line-of-business application" is a story about your work. Change the sector if the sector is the giveaway. Change the scale if the scale is. And say at the top of the page that details have been altered. Interviewers expect this. A candidate who has visibly sanitised their examples is a candidate who can be trusted with the interviewer's own network, and they notice.


Screenshots need the most care because they carry more than you meant to include. A dashboard screenshot shows the device names on every tile, the organisation's name in the browser tab, and your own username in the corner. Crop hard. Blur or overwrite what cropping cannot remove. If the graph is the point, consider redrawing it. The shape of a line and its two annotations are the evidence. A redrawn graph with a clean axis and fake hostnames proves the same thing with none of the risk. (It also looks better, which is not the point, but is not nothing.)


The write-up shape


Every artefact gets the same one-page write-up, in four parts. The parts are the problem, what you did, what the graphs showed, and what you would do differently. The fixed shape is a kindness to the reader, who can find the part they care about. It is also a discipline for you, because the fourth part is the one that proves you learned something, and it is the one people leave out.


The problem is two or three sentences: what was wrong or missing, who it affected, and how you found out. What you did is the design, the change, the runbook, in enough detail that a competent reader could repeat it and no more. What the graphs showed is the annotated image and a sentence per annotation: "the alert fired here; the dependency suppressed the other twelve; the line flattened when the SFP was replaced". What you would do differently is honest and specific. "I set the threshold by guessing and had to move it twice" is a better sentence than anything a certificate can say. That is because it is the sentence of someone who will not guess next time.


One page is the limit. Not "one page plus appendix". An interviewer has minutes. A manager reading your career record has fewer. And a page you cannot fit onto one page is a page you have not finished thinking about. The graph counts as part of the page. If the graph needs a second page, the graph is too busy, and so is the point you are making with it.


Where to keep it


Keep the portfolio somewhere private that you control and that does not belong to your employer. A folder on a personal machine, backed up. A single document with one section per artefact. A personal site if you want it to be findable, though most administrators are better served by a folder they can email a page from than by a site nobody visits. The choice of place matters less than two rules. The portfolio is not on the employer's systems, because it will not be yours when you leave. And it is backed up, because a portfolio lost with a laptop is two years of evidence gone.


Keep the unsanitised originals separately, if you keep them at all, and never in the portfolio folder. The portfolio is the thing you might one day attach to an email in a hurry. A folder that contains only sanitised pages cannot leak anything, however hurried you are. Treat the two folders as different countries with a border between them, and cross it only with the checklist below in hand.


Tip: add to the portfolio the same week the work happens, while the graph is still in the monitoring system and the surprise is fresh. Retention will roll the graph up into a smooth line by next quarter, and the smooth line does not show the spike that made the story. Capture it while it is jagged.


Using it in an interview


Offer one artefact when asked, not the whole folder at the start. Interviewers ask "tell me about a time" questions because they want a story with evidence in it, and the portfolio is designed to answer exactly that. When the question comes, answer it in two sentences from memory. Then say "I have a one-page write-up of that if it helps", and offer the page. Most interviewers will take it. Some will wave it away and ask you to keep talking, which is also fine, because you have just told them the page exists.


Walk them through the graph, not the text. The text they can read later. The graph, with you pointing at the annotations, is where an interviewer sees whether you understand what you are showing. That is the moment that separates a candidate who did the work from one who has a folder of someone else's. "This is where the uplink started dropping frames. This is when the alert fired, which was later than I wanted, and here is why. This is the replacement, and this is the flat line afterwards." Thirty seconds. Then stop and let them ask.


Bring paper. A printed page can be pushed across a table and left behind; a laptop screen has to be turned round and cannot. For a remote interview, have the pages open and offer to share your screen for one minute, and mean one minute. The portfolio is a tool for making one point vividly. A candidate who shows six pages when asked for one has stopped making the point and started making a presentation. (There is a time for presentations. It is after they hire you.)


The portfolio index


The index is one table at the front of the folder, one row per artefact. And it is the page you scan before an interview to decide which artefact answers which likely question. The sanitised column is a hard gate: a row with "no" in it does not leave the folder. The example rows show a portfolio that covers all six kinds.


ArtefactSkill it provesSanitised?One-line story
Design write-up: label-printing serviceDesigning checks and dependencies for a serviceYesThree checks and one dependency turned a weekly "host down" into a quiet service.
Dashboard: warehouse status wallBuilding for an audience; deciding what to leave offYes (redrawn)Eleven tiles, readable from the door; the twelfth space stays empty.
Runbook: disk space warningWriting for the person on callYesFollowed by two colleagues; one question, which became a new step.
Post-incident review: depot switch power lossLeading a review; reconstructing a timelineYes (organisation fictionalised)One alert instead of thirty; the action item was a second power feed.
Lab exercise log: seven exercisesPractising deliberately; noticing surprisesYes (lab only)Nine minutes is a long time to hold a cable.
Custom check: certificate expiryBuilding a check; documenting it for strangersYesWarns thirty days out; fails safe by reporting unknown, not OK.

Rebuild the index whenever an artefact is added or retired. Retire artefacts, too: a design write-up from your first year is worth keeping until you have a better one, and no longer. A portfolio that only ever grows becomes the folder nobody can find anything in, which is the same fate as most monitoring systems.


The sanitisation checklist


Every artefact passes every line before its index row says "yes".


  • Every hostname has been replaced with a descriptive name on example.com.
  • Every IP address has been replaced with an address from a documentation range.
  • Every username, email address, and personal name has been removed or replaced.
  • Every customer, supplier, and partner name has been removed.
  • The organisation's name does not appear, and the sector or scale has been changed if either would identify it.
  • Screenshots have been cropped, and anything cropping could not remove has been overwritten or the graph redrawn.
  • No credential, community string, shared secret, or key appears anywhere, including in code and comments.
  • No network diagram or address plan is included that would help an attacker.
  • The page states that details have been altered.
  • The sanitised copy has been read in full by a human, not only passed through a find-and-replace.
  • The unsanitised original is not in the portfolio folder.

One graph at Acme


A candidate interviewing at Acme for a network administrator role had a CV that said "responsible for monitoring across three sites". That was also what the previous two candidates' CVs had said. Asked about a time the monitoring had caught something before the users did, she took one printed page from a folder. It was a redrawn graph of an uplink's error counter climbing over eleven days, with three annotations. They were the alert threshold she had set, the day she had ordered the replacement optic, and the flat line after it was fitted. The interviewer, who ran the site's network, asked why the threshold was where it was. And she explained that she had first set it too high and moved it after the graph showed the errors starting earlier than the alert. They talked about that graph for fifteen minutes and did not return to the CV. She was hired. The graph is now pinned above the interviewer's desk as the example of what he wants new starters to be able to produce.


The graph did not prove that she was the best candidate. It proved that she had done the work and had understood it well enough to be wrong about the threshold and fix it. It proved she could explain it to someone who knew the subject. That is all a portfolio can do, and it is more than a CV can do in any number of pages. The two candidates before her may have been just as good. Nobody found out.


The portfolio checklist


Once a quarter, every line should be a yes.


  • The portfolio has at least two artefacts of different kinds, each with the four-part write-up on one page.
  • Every artefact has an annotated graph or screenshot with times on it.
  • Every artefact has passed the sanitisation checklist and the index says so.
  • The index is current and has a one-line story for each row.
  • The folder is on a machine I own, backed up, and separate from any originals.
  • I have added something in the last quarter, or retired something, or both.
  • I can answer "tell me about a time" for each row in two sentences without looking.
  • Each page can be printed, and I know which one I would hand across a table first.

Where to go next


A portfolio is six kinds of one-page evidence, sanitised until nothing in it belongs to anyone but you. It is kept somewhere you control, and offered one page at a time when the interviewer asks for a story. It answers the question a CV cannot, which is whether you did the work and understood it. A single annotated graph answers that better than any paragraph.


The portfolio is the evidence half of a case. The other half is the record of responsibility and trust that turning monitoring expertise into career capital describes. Where certificates sit alongside all this is in certifications: which kinds are worth your time. All of it is part of Monitoring Skills and Certifications for Career Growth.


Put this into practice with SysPeep Monitor

Every artefact in a portfolio starts as a graph, a dashboard, or an alert in a real monitoring system. SysPeep Monitor collects over SNMP, WMI, and SSH, keeps alert history, and draws live dashboards, network maps, and reports from one web console. So the annotated graph and the dashboard screenshot are there to capture when the work happens. The free trial on a lab network is a legitimate way to produce the lab exercise log and the custom check before the job does.

Download Free TrialLearn About SysPeep