South Africa

Builder · Catalyst · Executive

01 · Cover

Sybrand
De Kock

Complexity simplified.

Builder. Catalyst. Executive.Fifteen years turning broken systems into compounding ones, and the people standing next to them into operators who can do the same.

He prefers the complex problems, because that is where the most knowledge is gained.
Sybrand De Kock

South Africa

What follows

Receipts. Method. Philosophy. The record, and what it takes to reach him.

02 · The scene

The array was dead. So were the backups. Fifty-seven terabytes of a company's working memory, the indexed archive its clients paid to search, had gone dark inside a failed storage array, and the safety net that was supposed to catch it did not exist. A business that sold round-the-clock uptime was holding its breath. One person was put on the problem. He gave up the weekend and brought every byte back online.

Ask him why he chases that kind of work, and the answer comes without hesitation.

Saving companies from fire made me feel alive.

Most talented people optimise for the clean brief and the solved problem. He walks toward the version of a problem everyone else is quietly hoping to avoid, on his own initiative, when nobody is forcing him, because that is where he learns fastest and where the work is real. It costs him weekends, and it is the most useful thing about him.

03 · What you are looking at

The people who read this page can already hire the best specialists money can buy. They own rooms full of them. The disappointment tends to have the same shape: each specialist solves their one slice and goes blind to how the slices lace together, and the person who was supposed to see the whole board is nowhere in the building.

He is the one who sees the whole board. A generalist who gets sharper as the problem gets more tangled, who learns a new domain fast enough to out-reason the people who have lived in it for a decade, and who holds the machine and the boardroom in the same hand without a seam. Specialists you can buy by the dozen. The scarce hire is the one who watches every specialist defend their own square, then shows where the board is actually moving.

The five-word version is the one he says himself.

You dream it. He builds it.

04 · The engine

Solve it once,
for good.

One rule explains most of what follows, in his own words: I choose the hard path now, so it becomes simple later.

Most people take the easy route through a problem and pay for it afterward, in interest, forever. He takes the steep route going in, on purpose, because that is where the understanding lives. Do that for fifteen years and complex things stop looking complex. That is why he can make a thing simple for everyone else. He paid, up front, to understand it deeply enough to compress it.

The same discipline runs through unrelated domains, which is how you know it is a value system and not a fluke. He rebuilt an entire server room over one weekend, new hardware and a full platform cutover, because the quick patch was wrong and the rebuild was right. He runs penetration tests against his own production systems, hunting for the failures nobody asked him to find, before they surface the expensive way. He starts every job with the boring foundational work, the backups and the documentation and the observability, because that is the debt he refuses to leave for a future version of the system.

Underneath it sits a stubbornness that is almost temperamental. He keeps going at a problem long after most people have filed it under impossible, and often enough it turns out it never was.

05 · The range

Read the record straight through and the range is almost implausible for one person.

At sixteen he was deploying and updating Citrix environments across three hospitals on his school holidays. By thirty he had moved through retail repair counters, MSP field work, fibre trenches, server rooms, ransomware recoveries, production AI, open-source migration, and executive IT leadership. He has terminated cables on a Saturday and shipped an enterprise platform the same year. He has recovered dead RAID arrays and trained production AI. He has run fibre into residential developments and built a conversational AI agent from zero that now handles customer conversations at scale. He has sat in government tender meetings and written iOS forensics reports for law firms. Hardware at the rack, AI at the agent, security from the audit to the live test, infrastructure from the cable to the cloud.

One operator moves between all of those rooms and sees how they connect. The more tangled the problem gets, the more that instinct is worth.

06 · The proof

1.6M

AI conversations / yr

A production conversational AI agent he built and trained from zero. His own build, top to bottom.

85%

Customer satisfaction

The agent outperforms the human team it replaced

70+

Projects shipped

Infrastructure, security, automation. Under two years as Head of IT

300+

Tools migrated

Proprietary to open-source via Imbertech. Millions saved

57 TB

Recovered, zero loss

Dead array, no working backups

98%

Recovered

Total ransomware compromise. The company stayed open

National champion

Esports. Back-to-back League of Legends national titles at MSSA, 2016 and 2017

10+

Certifications

AWS, CompTIA A+, Network+, Linux+, Sophos, and more, on fifteen years of production work. Self-taught and credentialed both

07 · The catalyst

His best work
has a pulse.

The receipts above are what he built. The following is the part most builders never get to, and the part he treats as the actual job.

He does the work, and then he does the rarer thing. He makes everything around him more capable. He builds the system and trains the people until the whole operation runs at full speed on its own, and that durability is the value a company keeps long after the build is done. In his own words, and this is the line to keep:

I go further by lifting the people around me.

The evidence is people. He walked into an IT function that ran on reaction and firefighting and built a proper engineering department out of it, hiring, training, and handing real ownership to engineers who now carry large projects on their own name. He once coached someone from a working chef into a senior IT engineer, which tells you how far he will take a person who genuinely wants it. Right now, in his own time, he mentors more: two IT professionals, a senior software engineer, a junior support tech, and three people trying to break into the field. He calls those coaching days without a trace of irony, because the reciprocal investment is the point. At an earlier company he ran security-awareness training for the entire staff and kept a room of non-technical people genuinely paying attention with a bag of chocolates handed out for correct answers, which tells you he thinks about the human system with the same care he brings to the machine.

08 · The man

What drives him is thirst. He wants to know how things work, machines and people both, with the same hunger, and he goes after a person's psychology with the same seriousness he brings to a codebase. Money has never been what gets him out of bed.

That double appetite is why the range holds together. He learns a network the way he learns a room, by asking how it actually behaves instead of assuming, because the assumption bores him and the answer is the interesting part. It is also why the people who have worked closest to him describe the same experience: they got better, they got seen, and they got handed the keys.

And the curiosity runs a workbench. At home he keeps a full lab, Home Assistant and network-wide DNS filtering on his own hardware, a private mesh network tying a server, a GPU workstation, a laptop, and a phone into one system he drives from a single console. He listens to the radio spectrum with a software-defined receiver, tracking aircraft overhead and decoding the telemetry his own utility meters broadcast. He is wiring current-clamp sensors onto his mains and solar so he can watch his own power draw as a live waveform on a dashboard he built. He services his own cars down to the brakes, filters, and fluids, diagnoses them himself, and wired a self-written diagnostics logger into one so he could read every sensor while he drove it. He reads psychology and Stoic philosophy the way other people read the news, and when he needs a model for how to work he reaches for someone who actually lived it, Newton grinding through twenty years of failed experiments, rather than a motivational poster. His own framing for the whole habit is a line from Feynman: what he cannot build, he does not understand.

He started at a retail counter fixing walk-in PCs, learning to explain complicated things to people who did not care about the technical details, and he will tell you it is still the most useful skill he owns. The fire came later. The curiosity was always there first.

09 · How he thinks

How he thinks
about the work.

01

He builds teams the way he builds systems. People are subsystems with hidden state, blocked dependencies, and observability problems. The debugging discipline that clears a production outage clears a stuck engineer. Same loop, different substrate.

02

He measures everything that matters. Numbers cut through committee politics. A satisfaction score is harder to argue with than an opinion.

03

He plans for the worst case, because it arrives. Disaster recovery, ransomware playbooks, single points of failure tracked before they fail. When it happens, he is already three moves ahead.

10 · Currently

Right now,
he is

  • ShippingA production AI agent running daily under his ownership, built in-house rather than bought.
  • BuildingImbertech engagements across logistics, advertising, and telecoms, plus a public savings calculator anyone can run against their own software stack at imbertech.co.za.
  • DevelopingThe next intake of IT operators, from working professionals to people breaking into the field. Several in active mentorship on his own time, and Sundays are coaching days.
  • ResearchingA personal AI operating system he has been building for himself: an append-only event log with reducers, a contract-audit layer, and heartbeat timers, wiring an AI into roughly thirty custom integration servers and more than three hundred tools, with memory consolidation, pattern recognition, episodic recall, and causal reasoning. Fail-visible by design, any component that mounts but never reports back goes red in the audit. The most complex thing he has built.

11 · Next

The story is still being written. Do you want to be part of it?