Thought Behind Things · Mar 7, 2025
Pakistani businesses are one exploit away from going to zero
Muhammad Zayn — one of the youngest CISSP-certified professionals in Pakistan — walks through what cybersecurity actually means for small and medium enterprises, why on-premise data storage often beats the cloud on cost, and what really happens when a production house loses years of work to a corrupted firmware.
with Muhammad Zayn
11 min read
Why a content-agency owner went looking for a NAS
The episode opens with Muzamil framing the conversation against his own working problem. He has been watching Pakistan’s export-based services economy — freelancers, agencies, software shops — grow fast, and he has been running a content agency of his own. Scaling from twenty people to two hundred and fifty, he says, demands structures and tools that most Pakistani founders have never thought about. Data storage and data security are the ones he keeps tripping over.
That is what sent him looking for Synology, one of the top global data-storage and security hardware brands. He found it available in Pakistan, but quietly — through Al Madina Enterprise, the authorised distributor. He reached out, and they sent Muhammad Zayn, one of the youngest CISSP-certified professionals in the country. The promise of the conversation, as Muzamil sets it up, is a practical map of what enterprises, SMEs and serious freelancers can actually buy to protect themselves.
From graphic design to CISSP
Muzamil asks Zayn how he ended up in cybersecurity. The answer is not a straight line. Zayn went through a Karachi schooling track, studied computer science at college because the family expected him to land in an IT company, and along the way tried photography, content creation and graphic design — failing at the first two, getting traction with the third. COVID arrived. He landed an entry-level marketing role inside an IT-adjacent foundation, and from a desk meant for design work he started watching the network engineers.
“I was curious,” he says. “I went to my IT manager and I asked him, how are these routers communicating with these devices? How does this actually work?” He started studying for a Cisco CCNA through a welfare programme, then moved into networking and security properly. He cleared his CISSP through Al Nafi, a Pakistani institute that at the time was teaching the certification — normally a multi-thousand-dollar course globally — for around PKR 5,000 a year.
Muzamil pauses on the career arc itself, because he thinks it is the more important story for the audience. “Graphic designers are becoming redundant,” he says, “because of AI. You don’t need to hire a person for a logo any more. You can get it from an AI online.” Cybersecurity, by contrast, has the inverse curve. Every company that has a door has a guard outside it. Every company that has a server, he argues, will need a security person inside it.
Zayn agrees and sharpens the point. The Pakistani problem, he says, is not capability — it is awareness. “Leave the hacker,” he says. “I would be very interested in your data. I can build an alternate business out of it. You name it, I can do it with it.” Years of work, he reminds Muzamil, can be ground to zero in minutes.
The CIA triad, in plain words
When Muzamil pushes Zayn to define what cybersecurity actually is, Zayn does what every CISSP-holder does — he reaches for the CIA triad. But he translates it cleanly.
Confidentiality means only authorised people can access an asset. Integrity means even an authorised person cannot alter that asset without going through an authorised channel. Availability means the asset is reachable at the moment it is needed. Every storage and security decision, he says, maps back to one of those three letters.
He folds Muzamil’s specific worry — a fifty-terabyte media library spread across hard drives that occasionally fall and break — into the availability bucket. “We never rely on a single point of failure,” he says. The principle is identical to bringing a second cameraman: equipment fails, so you design around it. Inside a NAS that means RAID, drive-level redundancy, and the discipline to assume one or two disks will die in any given year.
What an actual on-premise storage build looks like
Muzamil hands Zayn a problem statement instead of a question: a thirty-person company that grew out of Fiverr freelancing, with no real workflow infrastructure, asking what to buy first. Zayn’s answer is structured.
A single NAS, he says, can solve most of it. Synology and QNAP are the two top-tier brands globally — “just like Apple and Samsung,” competing on the same ground — but Synology’s user experience is the one he keeps recommending, because a non-technical owner can configure it. There is no shell, no CLI, no coding. Everything sits in a GUI, and the mobile app controls the server.
The build, in Zayn’s telling, sequences cleanly. First the purpose: storage plus team sharing plus simultaneous project collaboration. Then the user count and topology — thirty people, mixed local and remote. Then the working data size. Then the location of the heaviest workload, because that is where the hardware physically goes, so the heaviest users hit it over a wired peer-to-peer connection instead of fighting Pakistan’s internet. Then access control lists, so an editor cannot see the accounts folder.
Then the collaboration layer. Synology Chat replaces WhatsApp for confidential internal communication, with the benefit that disabling a leaver’s user account closes the leak. Synology Office replaces Microsoft Office for shared documents. And the last layer is a heartbeat connection — a passive identical unit that takes over automatically if the primary fails. “The user sitting in front does not even know that a server has failed in the back,” Zayn says. “Within a minute or two, his work resumes.”
The on-premise versus cloud argument
Muzamil walks Zayn into the obvious objection. Why not just put all of it on Amazon? The cloud is the future. Hyperscalers have already solved redundancy, security, and global access.
Zayn’s counter is two-pronged. The first prong is reliability over the internet — in Pakistan, that is not a given. On-premise puts the working users on a direct wired connection to the server, with cloud-style remote access available as a hybrid option for travel and remote teams. The second prong is cost, and he is specific. He describes a Pakistani manufacturing firm with two-hundred-and-fifty to three-hundred users running Microsoft 365 for email. Their annual licence cost was around PKR six million — sixty lakh per year, recurring. The in-house replacement, deploying the same email functionality on their own server, came in at PKR fifty-eight lakh as a one-time hardware spend. Roughly one year of cloud licence fees bought the building.
He scales the same maths down. For a one-terabyte SME workload, the two-to-three-year cloud cost is roughly the same as a one-time in-house solution that delivers everything the cloud offers — multiple users, remote access, expandability — with no per-seat licence ceiling. “If I have a hundred-terabyte solution and a hundred people use it instead of thirty, there is no cost difference. You just optimise for performance.”
Concrete numbers for a real Pakistani buyer
Muzamil pushes for a ballpark. Fifty terabytes, working data, what does it cost?
Zayn does the arithmetic on air. Four sixteen-terabyte drives, merged with no redundancy in a basic four-bay Synology unit — the cheapest configuration he will quote — comes in at around PKR 700,000 for roughly fifty terabytes of usable space. Add redundancy at the drive level, and the figure moves to PKR 800,000 to 900,000 — still under a million. He is direct that the redundancy version is the only one he would actually recommend. “It is a considerable amount of data,” he says. “Drive-level redundancy is the minimum.”
He notes one small mercy of software RAID specifically. If the box itself dies — not the drives, the chassis — the drives can be pulled out, slotted into another Synology of the same OS family, and the array reassembles within ten minutes. The data does not need the original hardware to survive.
He scales the ceiling as well. The same hundred-terabyte solution can cost PKR 1.7 million, PKR 10 million, or PKR 100 million depending on drive type — spinning disk, SSD, NVMe — connectivity, controller redundancy, and high-availability pairing. The number is a function of how many failure modes the buyer wants to design around.
When the disaster is the building, not the hacker
Muzamil names the two Pakistani disasters that bend the rest of the conversation: load-shedding and an internet supply that is now being load-shed itself. Zayn answers with the disaster-recovery vocabulary that CISSP candidates are drilled on.
Power-supply redundancy is the first layer — dual power supplies, one on mains, one on UPS. Controller redundancy is the second — two controllers inside a single chassis so that pulling one does not stop the system. Site redundancy is the third — an identical off-site or hot-site unit, anywhere from a hundred metres to a kilometre away, running active-passive. The conversation widens to fires and natural disasters, which he treats as a normal CISSP topic. “CISSP will even cover how many types of fire hydrants there are, what type of fire could happen, chemical-based — all of that is in the syllabus.”
For remote teams over patchy internet, his preferred pattern is project-folder sync. The live project folder is synced to the laptop, the archive is not. A travelling editor has the live project on their machine even when the internet fails; their changes apply when connectivity returns. The wider archive stays on the central NAS.
What is already happening to Pakistani businesses
Muzamil presses for real Pakistani failure stories. Zayn gives one in full, with the names redacted. A Pakistani music production house was running storage with no second copy. An improper shutdown corrupted the firmware at the base level — not the drives themselves, the firmware. The data was technically still on the platters. None of it was reachable.
The shoot was pre-production, not post — models, sets, location work, crores of rupees of investment. The equipment vendor took no responsibility. Zayn’s framing is the line every CISSP candidate memorises and most business owners do not: data ownership and copy-keeping is the sole responsibility of the end user, the data owner. The documentation literally says so.
The recovery in this case worked. The principal escalated through the manufacturer’s technical team and the data was rebuilt. “But they took a one-million-dollar lesson from that,” Zayn says. “They keep multiple copies now.”
His broader stat is the one Muzamil pauses on. Global numbers, Zayn says, suggest that around ninety percent of organisations that take a serious data-loss incident shut down within two years. He is not citing it as a scare tactic — he is citing it as the reason cybersecurity belongs in a normal Pakistani SME budget, the same way insurance does.
He also clarifies the threat model. Ransomware exists. The breach where a production line gets locked until Bitcoin is paid exists. But the more common attacker is automated, not manual. “Every attack is not a person sitting at a screen targeting you. Scripts are running. Whatever they catch, they catch. The easier exploit is the best target.”
Immutable backups, deduplication, and snapshots
The final stretch of the conversation is the most technical, and Zayn translates as he goes. Synology’s enterprise tier includes what is called application-aware backup — the system can back up a running virtual machine without taking it down. Immutable backups mean the backup, once written, cannot be altered by ransomware that gets into the live environment; the attacker can encrypt the production data, but the backup snapshot is untouchable. Deduplication means the system inspects the data at block level, recognises duplicated blocks, and stores only one copy — so a hundred-terabyte raw workload compresses dramatically before it ever hits a disk.
The combination matters because of what it buys an SME owner. “He can recover from a snapshot five minutes ago,” Zayn says. The ransom demand becomes irrelevant. The downtime collapses from days or weeks to minutes.
By the end of the conversation, Muzamil is making the case to his own audience as much as to Zayn. The Pakistani SME — the one moving two or three hundred thousand dollars a month in services, the one whose entire output sits on twelve external hard drives stacked on a desk — is exactly the kind of business that needs a Zayn on retainer. Not a one-time install. A monthly check-in, an audit of who has access to what, a review of the metadata on last-access logs. The work is unglamorous and the saving is invisible until the day the firmware corrupts.
“You cannot take this very, very lightly,” Muzamil says, near the close. The conversation is the argument for taking it seriously while it is still cheap.
More from Thought Behind Things
Jun 20, 2026
The space economy's real wealth is in the startups under SpaceX
Muzamil reads the space-tech decade through one variable: the falling cost of reaching orbit. As that number drops, hundreds of companies and millions of jobs open up beneath the headline names.
Listen →
Jun 16, 2026
SpaceX's IPO is a pump. The space industry is real.
Muzamil reads the SpaceX IPO line by line: a 2 trillion dollar valuation on 18 billion in revenue and a 5 billion dollar loss, the index-fund rule that forces the buy, and why the real value is the hundred startups underneath.
Listen →
Jun 9, 2026
How Asad Mehmood landed Mattermost from Pakistan before A levels
with Asad Mehmood
Asad Mehmood walked into Mattermost before he had A levels, crossed two million dollars on Upwork, and now runs a design agency from Pakistan. He sat with Muzamil to lay out the framework underneath it: become undeniably good, then become visible, then sell outcomes.
Listen →Never miss what's next.
The dispatch - new writing and conversations, straight to your inbox.
First name, last name, email - in your inbox weekly. No spam.