How Asad Mehmood landed Mattermost from Pakistan before A levels

A guest who walked into Mattermost without a degree

Muzamil opens the conversation by framing why this episode exists. He has been running a continuous series on AI on Thought Behind Things, and last week’s podcast — to his surprise — crossed fifty-five to sixty thousand views. The audience signal was clear: people wanted the macro analysis grounded in the experience of operators actually doing the work. So he brought on Asad Mehmood, a Pakistani designer who currently runs his own UI/UX agency and works full-time at Mattermost from Pakistan.

The numbers are not abstract. Muzamil notes that Asad shared his documents before the recording — “Pakistan में यह बड़ा problem है. पहले हमने life में बड़ी सारी conversations की हैं. चूना बड़ा लगा है.” — and verified that Asad has done over two million dollars of revenue across his career, with more than a million of that on Upwork as an Expert-Vetted designer.

What makes the story land harder is the entry point. Asad explains he started working immediately after O Levels because his family’s financial situation did not permit a long career runway. “जब मेरे पास A levels भी नहीं था तब ही मेरी यह Mattermost की actual job लग गई थी,” he says — he was hired into a serious remote role before he had even sat A Levels.

Mattermost, NASA, CERN, and the credential excuse

Muzamil keeps the conversation granular. The first question he refuses to let pass is the one most Pakistani viewers will be silently asking: how is this possible without the right university on the CV?

Asad’s answer is structural. Mattermost is not an obscure shop. Its users include NASA and CERN. Its CEO previously worked at Microsoft. The colleagues Asad collaborates with daily are exactly the kind of people one would expect to be filtered through elite universities. “मैं Microsoft के लोगों के साथ काम कर रहा हूं. मेरे पास क्या था? A levels तकनीयता.”

His agency, he adds, currently employs people with Master’s degrees. The hierarchy that Pakistanis assume — credential first, opportunity second — does not hold once you cross into a market that is buying skill and presentation rather than transcripts. “If you actually have the skills and you know how to communicate and you know how to present your skills, that is sufficient to actually land really nice jobs.”

The two engines: undeniably good, then visible

The framework Asad has been building — and which Muzamil promises to link in the description — comes down to two engines.

Engine one: you have to be undeniably good. Engine two: you have to become visible. The order matters. Visibility without proof, Asad warns later in the conversation, just makes you a louder imposter.

Where most Pakistani talent stalls, he argues, is at engine one — and not for the reason people expect. The blocker is not skill. It is ambition. “Ambition भी नहीं है that we actually have to be one of the best in the industry.” Most people benchmark themselves against the colleague at the next desk, or against the other Pakistani freelancers on Fiverr. That is the wrong room.

Asad’s instruction is specific: study the global market, not the room you are in. If you are a designer, the benchmark is not the fifty Bangladeshi or Pakistani designers competing for the same gig on Fiverr. The benchmark is the studio behind the world’s best websites and applications. He names two — Augment and Pixel Point — as the calibre he watches. “मैं वह benchmark set करूंगा और उस benchmark को लेकर जाऊंगा.”

Fill your environment with great work until it feels normal

Once the benchmark is set, the next step is not a course. It is environmental.

Asad’s prescription is to saturate your inputs with the work you are trying to match. Open Twitter, open Instagram, but instead of reels and noise, follow the designers behind the studios you have decided to chase. Consume their output daily until your sense of “good” and “bad” recalibrates. “You can’t really go ahead and start creating them until you internalize these things.”

Step three is direct contact. Asad has personally reached out to Alex from Pixel Point and others. “मुझे कोई चीज़ समझ नहीं आती. मुझे इनसे पूछना होता है. मैं directly इनसे बात कर लेता हूं.” Muzamil seizes on this point and amplifies it. The mentor or hero you are studying has limited time — your job is to come prepared enough that the small slice of time they offer you returns maximum value. Connecting with the global community is, in Muzamil’s words, “a very, very useful piece of advice.”

There is a cultural sub-point embedded here that Muzamil pulls out explicitly. He references a YouTube comment from a recent episode where a 22-year-old Pakistani training to be an astronaut was mocked for his American accent. Muzamil’s response is impatient. The reflex to dismiss anyone who has shaped themselves toward a global standard — to laugh it off as posturing — is exactly the comfort-zone instinct that keeps Pakistani talent local. “The moment I feel like I know a lot मुझे पता लग रहा था कि अब तेरी learning journey रुक गई है.”

Become visible by being useful before being hired

The second engine — visibility — is where Asad gets most concrete about distribution.

He starts with open-source. Mattermost itself is an open-source project. Many of the people who eventually joined the company first showed up as contributors to the repository. Their work was visible. They understood the system. When a hiring need emerged, the company did not post a job and search through random applicants; it hired the people already contributing.

For non-developers, the equivalent is showing up where the decision-makers are and giving away the work. “First thing is I would actually build the evidence trail.” Identify where your buyers live — LinkedIn, Twitter, X. Then publish, comment, give feedback, fix things for free. “Even if it’s just like a single thing that you can add value to, just go ahead and do that for free.”

Asad’s own outbound pattern follows the rule. He does not pitch a thumbnail or a landing page redesign for money. He goes to a founder’s LinkedIn, finds their landing page, and writes them a free breakdown of what is hurting their conversion and why their churn rate looks the way it does. He is not trying to close that day. “I have to be seen multiple times. I have to develop credibility.” A week later, two weeks later, a month later, the founder already recognizes his name. That is when the conversation about a long-term contract becomes possible.

Muzamil’s reading of this is sharp. The dying pattern, he says, is the freelancer with a three-day-old profile, two posts, and a cold message claiming “we are a premium agency, we are changing the world.” There is no proof, no presence, no track. “Your digital presence is very, very important. People think it’s for content creators. It’s not. Everybody is looking at your digital footprint.”

Output is dying. Outcomes are what get paid.

The shift Asad and Muzamil keep circling is from output to outcome.

The old freelancer pitch was an output: I will build you a WordPress website. I will design you a thumbnail. The market for that has collapsed because Replit, Claude Code, and ChatGPT will produce the artifact at near-zero marginal cost. Muzamil names the cost-of-production deflation directly — it is the same point he made in the previous episode that drew this audience in.

The shift Asad has internalized is to sell the business result instead. He uses the same thumbnail example. A creator does not need a prettier thumbnail; they need more views, and views map to revenue. The pitch becomes: I will optimize your views, here is the experience trail. “The human is required for the outcome,” Muzamil says, “and I think that’s where the mentality shift is very much required.”

This is downstream of how Asad treats clients in the first place. He runs his agency on what he calls a customer-obsessed mindset — a phrase he says is ingrained at Mattermost. He treats clients as partners. He shares memes with them. He helps them with marketing problems that have nothing to do with his contracted scope. The point is not to be a vendor solving a ticket; it is to be the person they rely on. The result is that his agency keeps a small number of clients on multi-year contracts and does not need to scale headcount to be profitable.

Upwork has slowed because the buyers moved

Muzamil notes from the numbers Asad shared that his Upwork revenue has cooled in recent years and asks why directly. Asad does not flinch.

“पहले जो हम landing pages बनाते थे… इस तरह की jobs आना कम हो चुकी हैं and that is a reality that we have to face and that is a reality that’s going to keep on escalating.” Even with $1.8 million earned on Upwork, his honest read is that the platform under-pays relative to working outside it once you have a brand and proof. More importantly, his ideal customer profile has moved. “Series A और Series B के founders Upwork पर नहीं बैठे हुए.” They live on LinkedIn. So Asad’s focus has moved there.

The new shape of the work, in his agency’s recent pipeline, is revealing. Three of their recent clients had already built or partially built a product with AI tooling. What they needed was help launching it — and, in some cases, fixing it after the launch, because the AI-generated application had performance and security problems that real users immediately surfaced. “AI just cannot get you that far,” he says, when it comes to ship-ready software.

This is the wedge Asad is positioning his agency around. The market for building the product is commoditizing. The market for making AI-built products actually succeed — reliable, performant, secure, launched, retained — is opening. And that, by definition, is outcome work.

The next two to three years: demand destruction before demand creation

Muzamil closes with a question about scale. He shares that just before the recording, he had a mentorship call with someone weighing whether to take a US job or start a services business in Pakistan. His advice, given the person’s variables, was to take the job. The reason is the timing of the transition.

“There will be more demand destruction than demand creation,” Muzamil says of the next two to three years. The existing services model is automating away faster than the next wave of productivity work is materializing. There is a lag. Starting cold during a lag is risky.

Asad’s read is compatible but more individually optimistic. He tells the story of a colleague abroad who was let go and is locked into the identity of a specific specialization that is shrinking. He has urged them to transition. They will not. “They just don’t want to evolve. They just don’t want to look at the other side of things.” His message to Pakistani designers is the same: expand the avenue, learn Claude, learn to build workflows and systems, stop thinking of yourself as just a designer or just a developer.

Muzamil lands the episode on Javed Hasan’s term: creative destruction. The promised AI productivity gains require the old, inefficient systems to be broken first. A lot of people will get hit on the way through. “It is an exciting time and it is a terrifying time” — and the only real variable, in both his framing and Asad’s, is how quickly you decide to stop being rigid. Asad closes with Socrates, by way of the same point: “The only wisdom is in knowing that you know nothing.”