Pakistan's IT exports won't scale on cheap labour anymore

Why the startup conversation needed to move one floor up

The episode opens with Muzamil setting up a deliberate shift in framing. He has spent earlier conversations interrogating Pakistan’s startup scene head-on — the gold rush from 2019 through 2022, the collapse that followed, the question of whether the whole thing was a bubble. This time he wants to speak to someone who funds the layer above the startups themselves, and who can describe what investors are actually thinking now that the music has stopped.

His guest is Ahsan Jamil, the founding managing partner at Sy Ventures. Muzamil’s opening question is direct: is what we are seeing simply a burst bubble, or did the industry need to pivot?

Ahsan’s reply sets the tone for the rest of the conversation. “Startups are going through a correction,” he says. “A lot of the correction has happened.” He identifies a combination of forces — global capital constraints, local valuations that were unmoored from regional comparables, and a gold rush mindset that pulled in founders and capital without enough discipline. But the deeper diagnosis, the one that runs through the rest of the episode, is technical. The first wave was not technology-led enough.

The foundational technology problem

Muzamil presses Ahsan on the phrase “technology value addition” because, as he points out, the average user does not think in those terms. From the outside, the early Pakistani fintechs and food-delivery apps simply felt brain-dead — basic experiences that broke in obvious places. The one genuine local success story, he notes, was built by an outsider who quietly focused on the product experience rather than on marketing and then exited cleanly.

Ahsan’s explanation is precise. Most founders, he argues, treat the visible app as the product. They sweat the front-end experience. But “if the foundational technology underneath that is not architected and designed in the right way,” two failures compound. The first is that every release breaks something the user can feel. The second, and far worse, is that the company loses optionality. “You cannot enable features and you cannot grow and compete in the long run,” he says. Once a B2C product has ten million users and a brittle foundation, any meaningful change becomes a massive project.

This was the gap, in his telling, between the startups Sy Ventures backed and those it walked away from. The objection was rarely about valuations or multipliers. It was that “if you were going to beat your incumbent who’s got other advantages, you must be better than them in technology” — and most could not be.

Why Pakistani talent excels abroad but underperformed at home

Muzamil raises a personal version of the same question. He has brothers in big-tech roles abroad. Pakistanis routinely show up in Fortune 500 engineering organisations. So why, when that same pipeline of talent is exposed to the local startup industry, does the work come out weaker?

Ahsan’s answer separates need from capability. Locally, in the rush to capture market, the build approach became “very basic market capture point of view” — minimum viable products shipped quickly, with iteration used as cover for short-termism. “A lot of people use the word agile, a lot of people use an iterative approach as an excuse for just building things quickly,” he says. Globally, the same engineers operate inside organisations where the technology evolution is already several phases ahead, and the demand from that environment forces more ambitious architecture.

The cost of the short-term local approach is what Ahsan calls tech debt — the silent tax on every future iteration. The lesson he keeps returning to is that an iterative approach is fine, but only if the foundation is right. The first wave skipped that step.

The missing layer between technology and the real world

Muzamil sharpens the point further. He observes that a lot of Pakistani engineers, even very strong ones, sit inside services companies as back-office contributors. The architecture decisions, the client framing, the translation between a business problem and a technical structure — that work is done by someone else, abroad. Product management, in the sense of sitting at the intersection of real-world problem and technical solution, barely exists as a craft locally.

Ahsan agrees, and traces it to the type of work Pakistan’s services industry has historically taken on. “Nobody’s coming to you with a problem statement and saying, hey, I want to develop a personalisation platform that can scale to ten thousand transactions per second. Go help me.” The default mode has been headcount-for-hire — being bid into projects on cost rather than on problem-solving — and that is precisely the work that AI tooling is now eroding fastest. The companies in Sy Ventures’ portfolio that were doing “HR-ish” work in the beginning, he says, have had to restructure to go talk to global Fortune 1,000 clients and say “give us your problem; we will deal with the costing later.”

CodeNinja and the export-focused thesis

Muzamil asks for specifics, and Ahsan names CodeNinja as a portfolio company he can speak about publicly. The company was originally a services business with very strong aspirations to move from basic work into complex value addition. Today, he says, they work in artificial intelligence, cloud computing, and enterprise proof-of-concept solutions, primarily in the B2B export space.

This is the shape of the thesis. Sy Ventures invests in companies operating in frontier technology — AI, cloud, blockchain, complex data structures — that target Fortune 1,000 clients and earn revenue in dollars. The fund will tolerate a very large local client, like a major bank, but the orientation is outward.

The rationale is one of the episode’s most important threads. Ahsan is explicit about why he refused, on principle, to bring dollar capital into a rupee-denominated consumer play. “If you bring in dollar, let’s say, a hundred rupees and you’ve invested here, now dollar to rupee changes to three hundred and your consumer market for whatever you’re selling is local, that creates a lot of pressure.” That pressure, he argues, was a hidden driver of strategic risk-taking in the first wave — founders making high-risk bets because the return expectation had quietly switched on them.

His alternative inverts the structure. Local rupee capital, local talent, dollar-earning export revenue. The currency exposure stops being a threat and becomes either neutral or an advantage.

Five billion to fifty billion is not the hard part

Ahsan is unusually concrete about the size of the opportunity. Pakistan’s IT export exposure, he says, sits in the three-to-five billion dollar range depending on which slice you measure. Against the global AI and frontier-tech opportunity opening up over the next seven-to-eight years — which he sizes in the five-to-ten trillion dollar range — Pakistan is “a drop in the bucket.”

His framing is that this is good news. “For us to go from five to, say, fifty should not be that much of a challenge. We know companies which individually do that much business.” The economy only attracts serious competitive attention once it starts crossing five hundred billion to a trillion. Until then, the path from five to fifty billion is, in his words, almost a foregone conclusion “if we just do a few things right.”

Galvanising local capital and the limits of trust

Muzamil walks through the practical problem of where the capital comes from. Average Pakistani investors go into real estate files or, slightly more sophisticated, the stock market. There is no obvious on-ramp for them into the kind of fund Ahsan is running.

Ahsan is candid. Sy Ventures cannot today take in five hundred small investors because the institutional plumbing — the disclosure standards, the ongoing service expectation, the management overhead — is not there. He limits the LP base by check size and selection. The aspiration, he says, is eventually to put something into the public market so that “a person who is putting in five or ten lakh rupees” can buy units in a fund that is pegged against these companies. That day is not yet. When Muzamil asks whether the constraint is trust or ticket size, Ahsan replies that it is really a management problem — “it is easy to cater to the needs of twenty individuals and small institutions than it is to cater to two hundred” — sharpened by the fact that many Pakistani investors got burned in the first wave of startups and now hold the bar high.

A hybrid of venture and private equity, by design

Muzamil pushes on the business model itself. The companies Ahsan describes are already profitable. They are not Uber-style cash furnaces that need round after round of dilutive financing. So why is this venture capital and not private equity?

Ahsan is comfortable refusing the binary. “I don’t believe fundamentally in predefined notions of anything,” he says. The Sy Ventures model takes elements of both. The orientation is growth, in the venture style — even profitable portfolio companies are not paying dividends today because the capital is being reinvested to capture frontier-tech B2B market share. But the deal structures preserve optionality. Dividends remain available where they make sense. Exits remain available where they make sense. The pecking order, he emphasises, is the company first, then clients, then investors, then founders, then employees.

He is also blunt about Pakistan’s funding ladder. There is no realistic path here for a company to raise across a Series A, B, C, and D. The market is not deep enough. “You can have one round, one and a half rounds at max,” he tells founders. Anything beyond that has to be sold on a very specific multiplier story, because the buying liquidity for later rounds simply does not exist yet.

What investor capital actually pays for inside a profitable company

Muzamil presses on a question that often goes unanswered. If these portfolio companies are already profitable, where does the capital actually go? Office space and hardware do not absorb very much.

Ahsan’s answer is that the capital is a small part of what Sy Ventures brings. The real work is operational — institutional knowledge, client positioning, restructuring the company against global best practices. Engaging with Fortune 1,000 buyers requires runway: RFP costs, weeks or months of business development, proof-of-concept investments. But the deeper investment is inward. “How are you going to structure your people? What type of comprehension skills are you putting in place? What type of architecture skills are you putting in place?” The fund runs simulations so portfolio teams know how to engage with directors and above at Fortune 1,000 buyers. He is direct that this is not a marketing problem. “It’s not a marketing problem. Marketing is a part of the problem. The major problem is really structuring the organisation inwardly and its talent.”

Scale, the medium enterprise, and the AI-shrunk team

Muzamil raises a long-standing critique of Pakistan’s IT services industry: scale. India’s TCS can credibly walk into a Microsoft procurement conversation with two thousand people. Pakistan’s boutique software houses cannot.

Ahsan accepts the historic truth of the point — he tells a story about a 2023 Cisco conversation that required five hundred people, a bar Pakistani companies could not clear — but argues that the goalposts have moved. Two structural changes matter. First, Sy Ventures explicitly invests in growth and headcount where the work demands it; portfolio companies that need to collaborate on a client do so. Second, and more importantly, the size of team required to service even very large clients has collapsed. “Because of AI, the advent of all these tools in place, smaller, more nimble teams are able to service larger clients as long as they have the problem-solving mindset and the ability to integrate across systems.” The ratio of coders to architects, he notes, has flipped from something like ten-to-one toward roughly two-to-one. Pakistan does not have to win the Fortune 100. The Fortune 500-to-1,000 tier — where the thousandth company is still doing roughly fifty million dollars of revenue — is more than enough to absorb the kind of scaling Pakistan can credibly do over the next several years.

He also flags the American mid-market as an underrated opportunity. Medium enterprises in the United States, he argues, are unusually excited about AI-era partners because they can now compete with the giants in their categories without the capital spend that used to be required. “I am a small restaurant chain. I can compete with McDonald’s,” he offers as the customer’s framing. Pakistan’s portfolio companies, he says, are explicitly being pointed at that demand.

The diaspora, dollar capital, and the policy lattice around export businesses

Muzamil raises the obvious counter — a large Pakistani diaspora sits inside global tech, holds dollars, and would in principle be a natural investor base. Why is the fund not built around them?

Ahsan does not refuse diaspora dollars. The structure is set up so that dollar investors get dollar-in, dollar-out terms pegged against the dollar revenues of portfolio companies, which neutralises the currency risk they would otherwise worry about. But it is not the focal source of capital today because the local rupee capital base is enough for the current phase. The next phase, he says — aggressive international expansion of the portfolio — is where the diaspora network will start to matter more.

The conversation then moves to the policy surface. Portfolio companies have offices in the United States, the Middle East, and Germany. Capital movement out of Pakistan to fund that expansion is one of the friction points where policy still lags. Ahsan acknowledges incremental improvement — companies earning meaningfully in dollars can now reinvest a defined share abroad — and argues that the harder work is staying aligned with the policymakers on what trade-offs the macro can tolerate. The implicit point is that scaling IT exports is a policy project as much as a company-building one.

Why the macro picture does not need to consume every graduate

Muzamil closes the macro thread by asking the question that always haunts a Pakistani economic conversation: with millions of young people entering the workforce every year, can a strategy of capturing high-value B2B work ever absorb that population?

Ahsan refuses the framing. The age of AI has decoupled revenue growth from headcount growth, he argues. Infosys and TCS are restructuring globally for the same reason. “If you look at the layoffs that are happening in Microsoft and some of the bigger companies in the world, those layoffs are not happening at the expense of growth. They’re growing while laying people off.” The honest answer, he says, is that nobody yet knows how the population side of the equation resolves. But the only path to permanent wealth — wealth that survives rather than evaporating in the next cycle — is to do complex work at scale and let the second-order effects follow. “First capture this — twenty, fifty, hundred billion dollars from this side — then we’ll figure out what’s left.”

Tax, incentive design, and the case for a layered approach

Later in the discussion, Muzamil moves the conversation onto tax policy and the churn of salaried talent. Salaried professionals are over-taxed relative to retail, agriculture, and real estate, and the result is that a lot of skilled people are quietly arranging their lives so that the rupee economy sees as little of their income as possible.

Ahsan’s diagnosis is structural. The tax base does need to widen — the government, he allows, is working on that. But the relationship between Pakistan’s taxpayers and the state has become adverse because there is nothing visible in return. A CEO in the thirty-five or forty percent bracket gets effectively no individualised reward for participating. “What is the person who’s giving fifty thousand rupees in tax getting in return?”

Muzamil floats one idea — a blue passport for high taxpayers, easing visa friction for citizens whose mobility is otherwise constrained by Pakistan’s passport rank. Ahsan goes further. The right design, he argues, is not personal convenience but growth-oriented incentive layering: lower capital gains rates for taxpayers who invest in domestic equities and exit productively, structured rewards for behaviours that compound into the broader economy. “Convenience is the most simple tax return and tax benefit you can get. We need creative ideas that make people who are paying high taxes further participate in the economy and get a reward for creating upside.” The current policy conversation, he says, is not investing nearly enough brainpower in that direction.

A ten-year plan, not a fifty-year one

By the end of the conversation, Muzamil asks Ahsan for his image of Pakistan in 2050. Ahsan refuses the timeline. He references a public projection he has seen pegging Pakistan as a top-six economy by 2075 on the back of population alone and dismisses it as an aspiration too low and too late. “In the next ten years we should be in the top ten in terms of GDP and growth in the world.” The country is the fifth-largest population in the world, sits on a large talented youth base, and only needs to fix a handful of structural items. He cites neighbouring countries that have turned around in five-to-ten-year stretches when focused. “I want to see that when I’m still young enough to enjoy progress.”

Muzamil closes at the one-hour-seven-minute mark, asking listeners — particularly overseas Pakistanis — to engage with the talent question that runs underneath everything Ahsan has described. The conversation ends where it started: with the argument that Pakistan’s IT export future will not be won on labour costs, and that the institutions to win it on problem-solving still have to be built.