We recently connected with the founders of both companies to cut through the noise and take a hard look at how social enterprises are actually using AI on the ground and how it’s helping or hurting their attempts to scale. 

Their experiences reveal a striking paradox. On the one hand, AI is making social enterprises more effective, and that is, in turn, creating climate resilience and economic improvements for actual customers. At the same time, the gaps that have prevented poor and marginalized communities from accessing digital resources in the past (infrastructure, finances, language) are excluding them from the benefits of new AI-enabled services, threatening to leave the most vulnerable further behind. 

“We’re seeing a concentration of services going to people who already have services, whereas people who do not have services are getting left out even more.”

Aniket Doegar, Founder and CEO of Haqdarshak

AI lets humans do more 

At its best, AI makes humans more effective. When deployed strategically in companies with a social mission, it can create outsized impact.  

Aniket has seen this first hand with Haqdarshak. His social enterprise operates a network of 7,000 door-to-door agents, called “haqdarshaks,” who help Indian citizens and small businesses connect to state and federal government welfare schemes. For the haqdarshaks, AI-powered walkthroughs and probability engines make it faster and more effective to connect people with the benefits they have earned. Over the last 18 months Haqdarshak has almost halved their workforce while growing their household reach from 1M to 1.8M, almost a 4x increase in households reached per staff.  

That may raise eyebrows; it did for us. AI literally reduced the number of jobs. But it’s important to understand which jobs. Haqdarshak has not decreased its reliance on the human network of agents (its haqdarshaks) that do the actual work of connecting with people and then connecting them with benefits. What AI has done is make it much less intensive to manage that network, to monitor what agents are doing and where they are going. Similarly, it reduced the need for call center staff to follow-up on customers.  

Aniket told us, “We want to invest in building AI for our physical agents rather than replacing the physical agents.” What cannot be automated is the person who sits with another person, trying to understand them and their needs. That person cannot be replaced, but they can be made more powerful.  In fact, haqdarshak clearance rates have doubled, which directly translates to increased income.  

On Atram’s side, founder Davíd del Ser showed us how his company had built a team of AI agents to super-charge the climate readiness of lenders and their staff. Climate affects every aspect of a lending business: product design, screening, management, and collections. Having the right products, processes, and risk information in place — with a level of speed, accuracy, and customization unthinkable before AI — can help banks provide more climate-resilient financial services, years before a flood even strikes.  

The most forward-looking banks are already starting down this journey, and it is there that Atram is the most effective. “If you haven’t started your climate journey, call me when you do, because there’s no way we from outside can push you in the right direction. But the deeper you are into that journey, the more you love what we bring to the table”. Atram lets those companies design products that are more flexible and accessible, then bring them to bear in the places and moments they are needed.  

Concretely, Atram might encourage a bank to advise a shopkeeper in a flood-prone region to move their inventory onto shelving, or a farmer to harvest their berries before they are destroyed. These nudges can reduce the severity of impact; if paired with the right insurance or contingent loan product, they can facilitate rapid recovery.  

Interestingly, despite being an AI-native company (Atram started in 2024) humans remain fixed within its loop. The company layers multiple weather forecasts together, but it is a human meteorologist that ultimately makes the call to send out weather shock warnings. False positives are annoying, false negatives could be devastating.  

We connected with Karen Little, who leads Strategic Partnerships and Rapid Response at PayPal. PayPal has supported investment in Haqdarshak through Acumen and investment in Atram through Prosper Global Ventures. Karen told us that the pairing of AI and impact is something that PayPal sees across the small businesses that it serves. 

“Small businesses are the backbone of local economies, yet they and their customers are often the first to feel a climate or disaster shock and among the last to be reached by new technology,” she says. “At PayPal, we invest in entrepreneurs and in AI skills for small businesses because access reduces vulnerability and speeds recovery and long-term resilience. Both Haqdarshak and Atram show that AI works best when it extends the reach of trusted people and institutions. That is the kind of AI for small businesses we want to help build.” 

The infrastructure deficit 

The future is here, but it has never been so unevenly distributed.  

At its core, AI allows more trains, carrying more weight, to move faster than ever before on the digital rails of our economy. But that’s cold comfort to someone who is still pushing a cart up a hill. 

More than 2 billion people have still never been connected to the internet. In India alone, 600 million people (about 45% of the population) are offline. Those users are overwhelmingly rural and harder to reach. Doegar was honest about his frustrations with the gap between whom Haqdarshak wants to reach and whom it can profitably reach. 

“If I have to service a factory worker in Pune today,  I can do 50 services a day, and I can make a profit. The same service, if I have to deliver it in Uttar Pradesh [a populous but heavily rural state in northern India], my cost is 2.5x higher, my field person can do 10 services a week, and I don’t make any profit.” 

The barriers are numerous, and they are compounding. More rural areas are more likely to speak dialects that don’t have native LLMs. One of David’s biggest needs in West Africa is models that speak Yoruba and Hausa. They are less likely to be digitally literate or to own a smartphone. There will be less data on their borrowing and loss patterns. And unfortunately, from Atram’s perspective, they are less likely to be served by a financial institution that is digitally savvy and climate aware (digital literacy matters for B2B players too).  

These barriers are, del Ser told me, “losing us the people we most want to get.” They’re unlikely to be solved by a for-profit enterprise. But Doegar was excited about the idea of an ‘anti-AI fund’ that would make long-term bets on rural infrastructure. Gates Foundation and others are funding more work around real local language models. But we need even more funding on things like smartphone access, community support, and digital identity to ensure that people can benefit from AI advances, instead of being steamrolled by them.  

What social enterprises actually need 

The tools we are talking about enable outreach and customization at a scale that would literally not have been possible even two years ago. del Ser was excited to tell me: “We just talked to a thousand borrowers yesterday. That’s pretty cool.” But the pace of innovation cuts both ways. Aniket talked about how hard it was to stay focused when a new model drops every month; he could literally be retraining his teams at all times.  

When I asked both of them what social enterprises need the most when it comes to AI, del Ser was specific. There is loan-loss data within each financial service provider that would be an invaluable resource to predict future climate shocks and the depth of their effect. The tools to mine that data have never been better, but in many cases the doors remain locked. Creating more public datasets around climate-critical information (such as loss levels) would unlock smarter decision making.  

Doegar had a big-picture request. Given their customer base and public benefit, he sees a powerful case for social enterprises to access concessionally priced tokens. A similar concession was made years ago by AWS and Microsoft when it came to cloud computing, but the gap has re-opened now with AI. Lower-cost tokens and model access could create a boom of last-mile service models, or companies serving that market could get priced out. 

More from our Knowledge Hub

Insights All Regions

Beyond the lamp count

Could PULSE, a new Paris-aligned carbon credit methodology, unlock additional revenue for off-grid solar enterprises?

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Stories

Resilience, AI, and climate finance

Read more

Read more

The post The last-mile AI gap appeared first on Acumen.

First developed in 2010 under the United Nations’ CDM and revised several times since, AMS-III.AR helped kickstart the off-grid solar industry to great effect. However, AMS-III.AR suited a narrow group of large companies with the sales volumes to make the accounting worthwhile, and it was designed for a carbon market that has since changed considerably, both in what buyers expect in terms of credit integrity and in the way generated credits can interact with the evolving carbon market landscape.  

Now, a new carbon market is emerging under Article 6 of the Paris Agreement. To enable the OGS sector to take part, Gold Standard has published PULSE (Powering Universal Lighting via Solar Energy), a methodology that gives them a route to high-integrity, Paris-aligned carbon credits. PULSE is a direct output of technical assistance provided under Acumen’s Hardest-to-Reach initiative, to help portfolio distributors engage third-party carbon developers. Gold Standard’s published methodology acknowledges the support of Acumen and AGS Carbon in its development.

“PULSE isn’t a lighter version of the old methodology — it’s a more accountable one,” said Abhishek Goyal, a partner at AGS Carbon Advisory and a key contributor towards the development of the methodology. “We refined what counts and how it’s measured because the market it now serves, Article 6, expects nothing less.” 

The new methodology can be seen as a step forward, one that credits the light households actually receive and holds the sector to a higher standard. But we also recognize that greater rigor carries trade-offs. Here is our perspective on what PULSE changes, where we see opportunity, where we see risk, and what comes next.

What gets counted 

Until now, a company earned credits for the lamps it sold. The calculation was straightforward: Count the units sold, then apply a fixed emissions value for the kerosene lamp each unit replaced. That was the claim. PULSE keeps that route but adds another. It credits the light delivered instead. If a household system provides at least 1,000 lumen-hours of light per day, broadly aligned with the World Bank’s Tier 1 threshold, it can qualify through this new route. That threshold is equivalent to about five kerosene lamps’ worth of light, so a system that clears the bar is credited as replacing five lamps, whether it has one bulb or four. In other words, one well-made portable light can now count the same as five did under AMS-III.AR. 

PULSE also applies more deductions. Two are applied to the baseline itself: a fixed allowance for uncertainty and a country-specific adjustment that reduces the baseline over time, in line with national climate targets. A third covers so-called embodied emissions. Those from manufacturing the device, principally the battery and solar panel, are deducted from the credits it generates. In effect, the system’s own production footprint is set against the emissions it avoids. Two smaller deductions apply conditionally, one where displaced kerosene lamps are not collected and destroyed, and another where the device can also be charged from the mains. 

The net effect on credits per household is therefore not determined by the methodology alone. It depends on the system’s service level, the adjustment factor set for the country, and whether the company can demonstrate that the displaced lamps have been removed from use. There is no blanket sector-wide multiplier or figure. 

Where we see opportunity 

A credibility boost for a sector with real co-benefits. Renewable energy and off-grid carbon crediting has faced integrity criticism in recent years, and off-grid solar has sometimes been painted with the same brush despite being a different underlying case. A methodology built to Paris Agreement Alignment standards is a chance to make that distinction for the sector as a whole, rather than any single OGS company going it alone. The distinction matters for more than the carbon accounting. The sub-sector has traditionally targeted low-income households, and PULSE names the resulting co-benefits: reduced indoor air pollution, longer study hours for children, lower household spending on fuel. Those benefits can also have commercial value. Buyers have paid premiums for credits with documented development co-benefits.    

Carbon revenue as an affordability lever. PULSE states that its suppressed demand provisions, which credit the service a household should have had rather than the fuel it could afford to burn, are intended to direct carbon finance toward households whose fuel use has been low because they are poor. Where fuel spending was low because a household could not afford more, the accounting no longer treats that as fewer emissions to displace. 

Access to a different kind of buyer. Credits issued under PULSE are designed to be eligible for use in Article 6 transactions, subject to the host-country authorization. The significance lies in who the buyer becomes. Obligated demand should prove steadier than discretionary demand, which matters for distributors looking at revenue predictability rather than headline prices.   

Existing PAYG infrastructure can do double duty. Companies already running pay-as-you-go telemetry can use those logs as their monitoring record rather than building a separate survey program. For distributors that have invested in digital systems, a meaningful part of what PULSE requires is already in place. 

“Carbon finance has long struggled to reach the households most vulnerable to climate change,” said Susie Wheeldon, director of knowledge and insights at GOGLA. “PULSE recognizes that every family deserves enough power to study, work and stay safe after dark, and it credits projects on that basis. With digital tracking of every device and strong rules on battery and panel waste, it gives buyers the confidence to direct carbon finance to the communities that need it most. Unlocking that finance for off-grid solar can help lift millions more families out of extreme energy poverty.”   

Where we see risk 

The bar and the cost of clearing it. Device-level tracking, embodied emissions accounting, battery and panel waste management, and host-country authorization together make a demanding checklist, and every item costs time, money, or both. For a well-resourced distributor, this is a manageable cost of doing business; for a smaller or newer one, it may be out of reach. The risk is that access stays concentrated among the largest companies at the expense of smaller distributors often doing high-impact, last-mile work. PULSE anticipates this in part through simplified requirements in least developed countries and small island states, and through programs that let several companies share one monitoring effort.  

Authorization sits with governments, not companies. Article 6 host-country authorization is a sovereign decision taken on government timetables. Developers can build compliant projects and still find themselves waiting, sometimes indefinitely, on a signature that is not theirs to control. Early and transparent engagement with the ministries running a country’s Article 6 arrangements and its nationally determined contributions will be critical.² 

Carbon should not carry the business case. Credit prices are volatile and Article 6-market demand is still being tested in practice. PULSE also treats carbon income as temporary by design. At each crediting-period renewal, most activities must demonstrate that they still require carbon revenue to operate. Any distributor or investor treating such income as core, rather than a valuable but variable supplement, is taking on more risk than the instrument can currently support.  

Consensus is not the same as certainty. PULSE was built to earn broad agreement across a diverse set of stakeholders and largely achieves it. But questions will remain about the methodological choices underneath: where baselines are set, how far monitoring requirements should go before the cost outweighs the accuracy gained, and how any counterfactual can be established with confidence. Those debates are healthy, and the sector should expect them to continue.  

What happens next 

Could a Paris-aligned methodology open new doors for OGS enterprises? For companies able to meet its requirements, likely yes. New markets, more pricing power, and a stronger claim to legitimacy are all positive developments. For smaller and less-resourced distributors, that door may open more slowly, or only with additional support. What that support looks like is the open question: who builds and holds the monitoring systems, who carries the cost of running them, and how are the resulting revenues shared. That is the conversation we’re eager to have next.

More from our Knowledge Hub

Insights All Regions

The last-mile AI gap

How social enterprises are harnessing AI to scale impact — and where the tech falls short.

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Stories

Resilience, AI, and climate finance

Read more

Read more

The post Beyond the lamp count appeared first on Acumen.

Acumen’s perspective is specific and partial: 25 years of backing intrepid enterprises serving low-income communities that other investors deem too expensive to reach. That vantage point may look narrow but, to me, it is where three questions that affect all of us come into focus.

1. How do we price resilience in developing markets?

Resilience is easiest to understand through what people actually pay for. A farmer buys a solar irrigation pump because it turns one growing season into two. A trader buys a cold storage unit to reduce food spoilage. A business purchases insurance against a specific loss. They are buying a product or service that helps mitigate a setback, absorb a shock, or recover more quickly when one occurs.

For enterprises serving low-income customers, however, climate risk also shows up one step removed: through the financial resilience of the people they serve. A lost harvest cuts customer income, which cuts what a borrower can repay. And a single drought or flood can affect thousands of customers in the same place at the same time.

GOGLA surveyed 2,600 pay-as-you-go solar customers in Kenya and Uganda who were behind on their repayments. More than half said drought, crop disease, floods, or extreme heat had cut their income. Likewise, in India, increasingly frequent climate shocks are causing structural risk to the microfinance sector that serves over 70 million low-income households. In fiscal year 2024-25, portfolios at risk more than tripled from 2.1% to 6.2%. The weakening repayment capacity further drives up the cost of capital for the most exposed and least protected.

That correlation is what makes resilience difficult to price. Diversification works when losses are spread across customers, geographies, and time. Climate shocks can do the opposite, hitting whole communities and portfolios at once. In wealthier markets, mechanisms such as reinsurance, catastrophe bonds, and government backstops can help distribute that risk. In many lower-income markets, those protections remain limited or unavailable.

Resilience becomes investable only when entrepreneurs, lenders, and investors find better ways to price and share risk, not simply document exposure to it. At Climate Week NYC, I’ll be listening for examples of how investors are doing that — and what could translate to the markets Acumen knows best.

2. Will AI expand opportunity or concentrate it? 

Meaningful access to AI depends on two factors: an internet connection and reliable power. The map of who benefits from arguably the most significant technology shift since the internet is, more or less, the map of who already has that infrastructure in place. Without deliberate efforts to broaden access, AI risks compounding old inequalities. 

Capital flows point to the same challenge. Of the $42.7 billion in AI capital committed across five emerging market regions since January 2023, 94% went to compute infrastructure, rather than to businesses using AI to deliver products and services to customers on the ground. Infrastructure investment is necessary, but infrastructure alone does not determine who ultimately benefits. And yet, AI is accelerating what entrepreneurs can do. Crop2Cash, co-founded by Acumen fellow – Michael Ogundare, is an agricultural technology company that uses AI to provide real-time, personalized advice to smallholder farmers in their local languages. It is using AI to serve customers faster and cheaper to deliver at scale, not to make Crop2Cash an AI company.

We are seeing a similar dynamic in e-mobility, where AI can help companies optimize battery performance, manage charging networks, and operate fleets more efficiently. When those efficiencies bring down the cost of serving customers, cleaner electric 2- and 3-wheelers that were once out of reach can start to make economic sense for lower-income users.

AI can expand opportunity when businesses intentionally use it to lower the cost, improve the quality, or extend the reach of products designed for low-income customers. 

This has direct consequences for how Acumen invests. We are not looking for AI companies simply because they use AI. We are looking for companies where the technology can shift unit economics and create tangible value for customers who have historically been difficult to reach. As Climate Week NYC grapples with the AI revolution, that is the question I’ll keep asking: who benefits?

3. What actually moves money from commitment to action?

There is a hopeful theory that when wealthy countries feel the climate crisis directly, solidarity follows. Researchers have called this summer’s heat the most severe on record in Western Europe. If adaptation is becoming a line item in European budgets as a result, will that mean more funding for Africa and South Asia, or will it mean less? 

Developing countries will need more than $310 billion a year for adaptation by 2035, according to UNEP. International public adaptation finance is running at just $26 billion a year and that figure predates the steep aid cuts that followed. 

So what will it take to move capital into that gap?

Commitments are not enough. Capital moves when someone does the work of creating a market it can actually enter: matching different appetites for risk, solving for currency, and building the vehicles and pipeline investors need to deploy.

Match risk to the right capital. Our Hardest-to-Reach initiative brought 23 entities — philanthropists, impact investors, development finance institutions, a multilateral development bank, a global development fund, and a commercial bank — behind a single $250 million target, which closed in January. We built a dual, blended structure that allowed each to take a position they could actually hold. The point is not that every investor needs to take more risk. It is that the structure has to put the right risk in the right hands. 

Design for currency risk, not around it. The naira has lost roughly 70% of its value against the dollar since 2022. A company can be growing sales locally while still losing ground in dollar terms. That is not a side issue for international investors; it is part of the economics of investing in these markets. More local-currency debt can help. So can backing businesses whose revenues and costs are naturally aligned in the same currency. But where foreign capital is required, currency risk also needs to be explicitly priced, hedged, or shared rather than left for the company to absorb. 

Build the market, not just the pledge. Even deeply committed investors can struggle to deploy capital into climate-exposed markets if there is no fund or instrument through which to invest: no investable pipeline of companies, no local team to originate or assess opportunities, or no catalytic layer able to take risks that commercial capital cannot yet hold. Those pieces do not appear because a commitment has been announced. They have to be built, often over years.

That is the less visible work of climate finance, but it may be the most important. Market creation is not simply about finding more money. It is about building the structures that allow different kinds of money to move.

What I’m bringing to New York

In New York, many rooms will understandably focus on the mounting toll climate change is taking in the U.S. and Europe, the AI revolution, and the geopolitical forces reshaping global finance. My attention will also be on what those debates mean for the world’s most climate-vulnerable markets.

The thread connecting these three questions is design. Without structures intentionally designed to reach the people markets have often overlooked, resilience stays uninvestable, AI’s benefits stay unequal, and capital stays committed rather than deployed. The technologies, entrepreneurs, and business models already exist. The question I’m bringing to New York is whether we are willing to design the capital to meet their needs.

More from our Knowledge Hub

Insights All Regions

The last-mile AI gap

How social enterprises are harnessing AI to scale impact — and where the tech falls short.

Read more

Insights

Beyond the lamp count

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Read more

The post Resilience, AI, and climate finance appeared first on Acumen.

Despite that, impact investors, Acumen included, have not captured these patterns consistently, nor have we fully acted on the opportunities they point to. This piece sets out what de-risks a deal, what creates friction for the companies we back, and how those companies have built bridges to the policy environment. We’re sharing it in the hope of seeing greater alignment between public initiatives and private innovation.

What de-risks a deal

A few patterns reliably improve our portfolio team’s confidence in a deal. Others just as reliably erode it. Taken together, they sit on a continuum that runs from enabling to prohibitive. 

On the enabling side, predictable, product-specific regulatory policy stands out. Agroeknor, one of our Nigerian portfolio companies, owns one of only seven nationally accredited fumigation chambers in the country. That credential is hard to get, internationally recognized, and stable. It shows how a business model can be built around a specific enabling policy that has broad political support and durable funding.

Compliance regimes can work the same way. In cocoa, buyer and destination-market requirements — full traceability back to the farm, and verified freedom from child labor in the supply chain — are demanding, but they are also specific, documented, and enforced. A company that can meet them holds a credential its competitors don’t, and can price accordingly.

Government programs with clear eligibility criteria are another. When we reviewed Omia, an agri-input company operating in northern Uganda’s refugee-hosting districts, the rationale for expanding into a new district rested partly on Operation Wealth Creation, a national agriculture support program that had already put input-distribution and training infrastructure on the ground. 

Trade policy, when it’s clear, can be powerful. In Nigeria, a 2020 presidential directive restricting foreign-exchange access for food imports reshaped the competitive landscape for two of our grain-sector investees, Zebra Crop Bank and Cropsafe. Local aggregators suddenly had structural demand that didn’t exist when cheap imports flowed freely. In Uganda, our coffee exporter Mountain Harvest benefits from restrictive export-licensing regimes in neighboring countries that Uganda doesn’t impose. When our investment teams can point to a durable trade policy and say, “This is why a market exists for this company,” the selection case gets stronger.

What creates friction

The prohibitive side of the continuum gets discussed less, and may be more useful. It comes back to two themes. 

Policy ambiguity is the single greatest source of government-related friction in our due diligence. In one East African market, a proposal to fold a sector-specific regulator into a larger ministry was flagged as a standalone risk — not because the rules had changed, but because they might. Acumen can underwrite strictness. Neither we nor the companies we invest in can underwrite an absence of rules.

The second theme is the gap between import policy and export policy. Governments across our portfolio countries pursue import substitution, restricting selected imports to benefit local producers. What doesn’t always follow is the export-enablement infrastructure that lets those same producers grow beyond the domestic market. To export, a company may need fumigation accreditation, phytosanitary approvals, and an export license, each with its own agency, timeline, and fees. At the destination end sit tariffs that penalize processed goods. In between sits a working-capital cycle that runs 90 to 105-plus days from farmer payment to buyer wire, including roughly 60 days of sea freight. Stacked together, these layers systematically disadvantage small and medium-sized companies. One of our West African investees doesn’t export directly at all, despite strong international demand, because its balance sheet can’t absorb 30-plus-day receivables.

The power of public-private partnerships

In May 2023, we invested in Zebra CropBank, a Nigerian company deploying decentralized storage hubs at farmgates across the grain belt. The investment memo named Benue State as an expansion target and listed “state and local government partnership agreements” as a strategic priority. The founder, Buffy Okeke-Ojiudu, worked as a program management officer at the Economic Community of West African States (ECOWAS) — a major regional trade union — and built Entrepreneurs for Agriculture Transformation (E.A.T), a convening platform connecting agro-entrepreneurs with policymakers.

Within two years, CropBank signed a formal memorandum of understanding (MOU) with the Benue State Government covering land allocation, expedited regulatory approvals, and prioritized road and power infrastructure, in a process that also included the UN Development Programme. A year on from signing the MOU, the company operates across seven Nigerian states and into Cameroon, with more than 42,000 farmers engaged. 

We see versions of this across the portfolio: a Kenyan founder who built an industry association to lobby on import tariffs, a Ghanaian company with a ministry official as a board observer, a Sierra Leonean cocoa company petitioning for a free economic zone. None were structured as partnerships at the time of investment. All function as formal operating partnerships now.

The catch is that each of these partnerships exists because individual founders built it from scratch. There is no repeatable pathway, no standard government intake process. The Benue MOU shows what a structured partnership looks like when it works. The open question is how that model reaches founders who don’t have ECOWAS on their CV.

What we propose

For national governments: 

• Create a standardized pathway for agribusinesses to propose public-private partnerships, specifically for infrastructure co-investment — land allocation, regulatory fast-tracking, shared maintenance — rather than requiring each founder to broker their own MOU. 
• Publish policy pathways even when they’re incomplete. A draft framework with a timeline does more for investor confidence than silence.

For development finance institutions and bilateral funders: 

• Design instruments with explicit continuity language for what happens to portfolio companies if the funder’s own policy environment shifts. 
• Where governments won’t, fund the regulatory enabling infrastructure that rarely attracts attention: accreditation labs, produce-monitoring boards, export-readiness programs.

For fellow impact investors: 

• Build government and regulatory engagement, not just government and regulatory risk, into standard diligence templates, tracked with the same rigor as commercial and technology risk. 
• Treat a founder’s policy engagement as an asset to be assessed and supported, not just a line on a CV.
• Share patterns like these across the industry. Individually, each of us sees one country at a time, one deal at a time. Aggregated, the signal becomes actionable.

What comes next

While policy is almost always referenced in a broad, contextual sense, there’s far less discussion about the consistent themes that can supercharge or poison a deal. The frictions we’ve named are always in our risk assessment, priced into our deal structures, carried forward in our monitoring. Those are real costs borne by real companies, and eventually by the farmers behind them. 

The models for doing this differently are emerging. Work led by GAIN, FAO, and other partners through the Agrifood Systems Accelerator points toward new attempts at building the discourse and blending mechanisms that enhance alignment between sovereign commitments and private operational capacity.

If you’re an investor who recognizes these dynamics, we’d be interested to hear whether your evidence confirms, complicates, or contradicts ours. If you’re a government official designing agricultural policy, we’d welcome the chance to compare notes. The evidence our sector already generates should travel further than it currently does.

Support for this piece and Acumen’s Trellis initiative was provided by the Foreign, Commonwealth & Development Office (FCDO). 

More from our Knowledge Hub

Insights All Regions

The last-mile AI gap

How social enterprises are harnessing AI to scale impact — and where the tech falls short.

Read more

Insights

Beyond the lamp count

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Read more

The post Policy in the portfolio: A view from inside the deal appeared first on Acumen.

Walking back into a founder room in Skardu, after spending years of my life building social enterprises in Pakistan, I felt both at once. Familiar, because I’ve sat where they were sitting. Different, because for the first time, I was watching the dance floor from above.

It felt like coming home, the kind of familiarity where you catch yourself thinking, “These are my people.” Not just people solving similar problems, but people who understand the specific weight of trying to solve them in Pakistan — with all the complexity, urgency, and possibility that comes with that work.

The Skardu Summit Fellowship brought together 10 of Pakistan’s top social entrepreneurs to pressure-test how their models would hold up under growth, using the Mulago Foundation‘s framework. It was organized by the educational technology company Taleemabad, led by Acumen Fellow and Acumen Angels awardee Haroon Yasin, and supported by the Mulago Foundation, which actively scouts and funds high-impact, scalable organizations in Pakistan. The faculty in the room reflected the full ecosystem a founder actually needs: funders, seasoned builders, and government officials, sitting alongside each other, asking hard questions, and engaging in wide-ranging conversations about creating meaningful impact. 

The location was intentional. Skardu, nestled in the mountainous Gilgit-Baltistan region of northern Pakistan, drops you into a landscape built on a different scale. Mountains on every side, rivers cutting through valleys, and some of the highest peaks in the world at your doorstep. There’s something about that vastness that gets into your thinking and pushes you to think bigger.

The week was framed around scale. But for me, it also became a more personal reflection on what it takes to move from building an organization to building something that can truly meet the size of the problem you set out to solve.

Scale is not the same as growth

This is the line I keep coming back to: Scale is not the same as growth.

Growth is more revenue, more users, more headcount. Scale is your biggest bet on how you actually solve the problem you started with and everything it will take to make that bet real. As the Mulago framework puts it, it means finding something good enough, cheap enough, big enough, and simple enough that a doer will do it and a payer will pay for it. And it means iterating relentlessly until you have real evidence of impact and unit economics that hold up. 

But beyond the language of any one framework, the deeper question is this: What actually works, who needs it, and who else must carry it forward for the solution to reach the level of the problem?

During my time as COO at Taleemabad, we were obsessed with getting more children to reach grade-level learning across Pakistan. That’s what led us to partner with the government as the path to scale. We went through rounds of learning from low-cost private school models, stripping our model down, and simplifying it until we landed on a large government pilot. Eventually, the government became a paying customer, and today the model reaches over 300,000 students.

The lesson was not that government is always the answer. It was that scale did not come from doing more ourselves. It came from understanding the system we were operating in, identifying partners who could take it to millions of children, and asking who could carry the work further than we could alone.

The hard part isn’t having conviction about the problem. It’s learning from the data, making the hard calls to kill what isn’t working, and putting everything behind what is working — until your model is refined enough to scale.

This isn’t an easy process. A few founders hit the same wall that week in Skardu. They were doing too much. The sessions pushed them to dissect their models and find their biggest bet among all their activities, looking at the data to see what showed early evidence, what would be easy for someone else to carry out, and what could create an outsized impact. What fit their mission, and what were they doing just to keep the organization running?

Part of the breakthrough was simply picking what to focus on and letting the rest go. But the bigger realization was that scale isn’t something you do alone. It means working with others, learning from people who’ve done it or are a few steps ahead, so you can see what you’re missing. That is where community becomes more than support. It becomes infrastructure.

Community is scaling infrastructure

I have seen this in my own leadership journey as an Acumen Fellow. Through the right tools, skills, and peers around me, I could finally see where I needed to grow as a leader and where I was getting in my own way. The community did more than encourage me. It challenged me. It held up a mirror. It helped me step off the dance floor long enough to see the system around me, and my own role within it, more clearly. 

For founders working on problems of poverty, that kind of community is not a soft benefit; it shapes decisions. It helps leaders stay honest about what is working and what is not, and it gives founders the courage to keep going without confusing endurance with doing everything alone.

In Skardu, I saw that community again. Haroon, an Acumen Fellow, helped bring the room together. Acumen Fellows were there not only as participants, but as peers and guides, carrying lessons from their own journeys into conversations with founders facing similar questions. 

It was a reminder that Acumen’s work in Pakistan is not only visible through programs, investments, milestones, or application cycles. It lives in the people who keep building, keep learning, and keep showing up for one another. Acumen has been backing builders in Pakistan for over two decades. 

Builders like Osama Shahid, Founder and CEO of Soby Agro, an Acumen Fellow, Acumen Angels awardee, and Acumen Agri Accelerator participant. Osama has spent years building affordable farm machinery for Pakistan’s smallholder farmers — many overlooked by traditional providers because their farms are considered too small. Today, Soby Agro serves more than 10,000 farmers, dairy operators, and food processors with equipment better suited to local realities. In Skardu, Osama brought those hard-won lessons back to other builders facing their own questions of growth. 

The week in Skardu reinforced something I keep relearning: Community, capital, and capacity are critical inputs that make scale possible. That’s what I understood differently, watching from the balcony. Scale is never about how far you can go alone. The important question to sit with is this: How do you build a model that others can take to solve the problem at the size it actually exists?

More from our Knowledge Hub

Insights All Regions

The last-mile AI gap

How social enterprises are harnessing AI to scale impact — and where the tech falls short.

Read more

Insights

Beyond the lamp count

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Read more

The post What I understood differently after leaving the builder’s seat appeared first on Acumen.

It was all a bit surreal. I was sitting with 15 Acumen Fellows, who are building enterprises that solve some of the toughest problems on the planet. They work in the margins with communities that are still contending with the basics like access to healthy food, steady electricity, and dignified work. We’ve brought a cohort of our Fellows to TED every year for five years now. It’s inspiring every time, but never has it felt as important as this year.

The theme for this year’s conference was “All of Us” — the idea that navigating this terrifying, hopeful moment will require everyone: drone manufacturers and Acumen Fellows, AI companies and Nobel laureates like Malala Yousafzai, who opened the conference with a reminder that for millions of girls, the right to simply go to school or play a sport remains unfinished business. One speaker cited a phrase drawn from an OpenAI white paper making the case for data center investment. Infrastructure is destiny. I kept turning it over all week.

Infrastructure is destiny.

Not roads and railways, although those matter. The invisible kind of infrastructure. The algorithms that channel our attention and spending online. The drones that circle the skies without us ever knowing. The platforms, protocols, and code written by corporations that now wield the power of empires. The systems that pre-determine the shape of what’s possible before any of us sit down to make a choice.

TED2026 arrived at a strange and urgent moment. Every breakthrough at the conference came with a shadow. Drones for safety — and for surveillance. A billion biological experiments in the next four years to map RNA the way LLMs cracked language — and the civil rights questions that trail quietly behind. Agents that can now do everything a person can do on a computer — and the obvious question that raises about human agency in an agentic future.

One of the most powerful talks came from Carissa Veliz, who asserted that predictions are not facts, even though our markets and therefore our world treat them as such. Facts belong to the past. And yet we are living through a moment when the most powerful tools are being used to make the future feel inevitable — to close the conversation before it begins. The story being told about AI, about displacement, about who will benefit and who will be left behind, is not yet written. It is a choice being made, right now.

When I walk into TED with Acumen Fellows, something shifts. Their presence is its own kind of moral compass. Their presence cannot help but make you think about who is not there, because their life’s work exists within the gap between the world being imagined and the world they are building in — that gap is exactly where the moral infrastructure either gets built or doesn’t.

Niharika Jain, the founder of Broomees in India, is building the infrastructure of dignified work for domestic workers, training, verifying, and professionalizing a workforce that has historically been invisible to the formal economy. When AI is being deployed to automate the so-called “menial,” Niharika is asking the prior question: How do we make this work visible? How do we make it count? Dignity, she understands, is not a feature you add later. It is the foundation.

Willie Ng, the founder of Global Cerah in Malaysia, is turning agricultural waste into animal feed and fertilizer, reducing input costs by 40%, building a circular economy from the margins for farmers who have been traditionally excluded. He left an eight-year corporate career because he saw what was being wasted, and what was possible. His work is infrastructure in the most literal sense: the kind that feeds people, restores land, and closes loops that extractive systems leave open.

Mainak Roy, the co-founder of Simple Education Foundation in India, has spent more than a decade building the infrastructure of learning for children in government schools, training teachers, transforming classrooms, working toward 10 million children across the country. His unyielding belief guides him: one good teacher can transform a life He understands that  education is not a service, itis the infrastructure of agency itself.

These three — and the thousands of Acumen Fellows like them across our global community — are the human infrastructure. They are the steel and girders of a different kind of future. They have the expertise to understand the systems they are working within, the moral imagination to see beyond them, and the stubborn, grounded belief that the people most affected by broken systems are also the ones most capable of fixing them.

“Abundance without participation feels like scarcity,” Van Jones said. The infrastructure of opportunity — who gets access, who gets to build, whose knowledge is encoded and whose is ignored — does not assemble itself. It is built, or it isn’t, through deliberate choices made by people with the vision and the will to make them.

The razor’s edge between hope and fear is real. I felt it at TED2026 more acutely than any year before. But I also felt something else: the quiet, resolute presence of builders who have looked at the world as it is and chosen to build the world as it could be. A growing force that is not succumbing to the power of prediction or presumed “inevitability” of how the world will shake out, but rather actively shaping it through grit, will power and community, and a belief that we will be okay, if we ground in our shared humanity.

More from our Knowledge Hub

Insights All Regions

The last-mile AI gap

How social enterprises are harnessing AI to scale impact — and where the tech falls short.

Read more

Insights

Beyond the lamp count

Read more

Stories

Cuando la miel construye paz: el poder transformador de la asociatividad

Read more

Stories

When honey helps rebuild peace: The transformative power of collective action

Read more

Read more

The post Infrastructure is destiny appeared first on Acumen.