Why Is No One Building the Software Stack for Energy in Africa?

wondering if I'm the only one seeing this

UPDATE: As I was editing this piece, two massive things happened:

1. Nigeria just introduced a 5% fuel tax on petrol and diesel

2. NERC dropped draft Net Billing Regulations (50kWp to 5MWp)

If you needed a sign that NOW is the time, this is it.

---

Last week, I watched a brilliant Nigerian developer demo their new AI coding assistant. Beautiful UI. Impressive context awareness. Already had 200 users.

I asked them one question: "Have you considered building for energy instead?"

They looked at me like I had suggested they pivot to farming.

Meanwhile, I am sitting here watching the biggest infrastructure gold rush in African history unfold, and everyone's too busy building the next ChatGPT wrapper to notice. The developers who could be building Africa's energy future are instead optimizing React components for Silicon Valley's latest AI startup.

Let me show you what they're missing.

The $11 Trillion Question

Here's what keeps me up at night: Nigeria is about to deploy more distributed energy infrastructure in the next decade than it built in the previous century. The C&I boom is real. $750million DARES (Distributed Access through Renewable Energy Sources) is happening. Alternative energy financing is exploding.

And the entire ecosystem is running on WhatsApp and Excel.

I'm not exaggerating. I'm watching the entire solar value chain operate like it's 1995:

• Project Development & Origination? Manual. Completely manual. And there's no other way. Not because we're backwards. Because the data literally doesn't exist.

  • Google Maps? Still showing 2022 images. That empty lot? It's been a shopping complex for two years. That "residential building"? It's now a bank with 50 ACs running.

  • Customer energy data? "We buy 3,000 liters of diesel weekly." That's it. That's the data. No kWh. No load profiles. No power factor. Just diesel receipts and generator run hours scribbled in a notebook.

  • Grid consumption data? DisCos do estimated billing. Your "10,000 kWh monthly consumption" could be 5,000 or 15,000. Nobody knows. The meters don't exist, or they stopped working in 2019.

  So yes, someone MUST drive three hours through Lagos traffic to count ACs, photograph the actual roof (not the 2022 version), check if those are 1.5HP or 2HP units, see if the generator is 100kVA or 150kVA, and ask the facility manager how many hours they actually run it.

  This isn't inefficiency. This is the reality of building infrastructure where data infrastructure doesn't exist.

  Which makes the software opportunity even bigger. We don't just need project management software. We need to BUILD THE DATA LAYER that doesn't exist. Create the energy consumption baseline. Establish the load profiles. Generate the data that makes everything else possible.

• Licensing? Print the forms. Drive to NERC and FMEnv. Wait in line. Get a stamp. Pray nothing changes while you wait 3-6 months for approval.

• Power Purchase Agreements? Some use DocuSign (progress!). Most? Email Word docs back and forth until someone prints it, spiral-binds it, and drives it to the Chairman's house for signature. I've seen $5 million deals closed with documents that nobody can search because they're scanned PDFs of photocopies.

• Design & Engineering? Sure, there's software. American software. European software. Software that thinks every roof is flat and every day has 12 hours of sunlight. Software that doesn't know what harmattan is or why you need to derate by 30% when it's 45°C in the shade.

• Financing? "Innovation" so far has meant a fintech with a solar hobby (hello PAYGos). And here's the thing - finance NEEDS scale to work, but our current tools make scale impossible.

  • Try raising a $50 million fund when your pipeline is tracked in WhatsApp messages.

  • Try convincing DFIs to invest when your due diligence is 47 different Excel files with broken formulas.

  • Try aggregating 100 small projects for institutional capital when each one has different models, different assumptions, and different ways of calculating the same damn IRR.

The money exists. Climate funds, DFIs, impact investors - they're sitting on billions looking for African renewable projects. But they need:

• Standardized data rooms (not Google Drives with 1,000 unnamed folders)

• Aggregatable project pipelines (not individual Excel files)

• Risk assessment at portfolio level (not WhatsApp voice notes about each site)

• Track records provable by data (not "trust me bro, we built 50 projects")

Instead, what do we have?

• Financial models with hardcoded exchange rates from 2019

• Due diligence via email chains nobody can follow

• Investment committees making decisions based on PowerPoints with made-up utilization rates

• No way to aggregate small projects into fundable portfolios

• No standardized performance data to prove track records

The cruel irony? Solar needs to be deployed at massive scale to hit climate goals. But it's being financed one painful project at a time because we don't have the software infrastructure to aggregate, standardize, and scale.

This isn't just inefficient. It's why African renewable projects pay 15% for capital while European projects pay 3%. The "Africa risk premium" is really just a "shitty data infrastructure" premium.

• Installation & Construction? Project management via WhatsApp groups named "Site 1 - Lagos Solar 50kW (DON'T DELETE)." Daily progress? Photos in the group chat. Quality control? "Bros, panels look correct?"

• Commissioning? Someone drives to site. Takes photos. Sends to WhatsApp. "System is working ✅"

• Operations & Maintenance? One guy. On site. With a smartphone. Sending blurry photos to a WhatsApp group whenever something breaks. "Monitoring" means checking if the inverter's LED is green. Performance analytics? "It was working yesterday."

• Real-time Monitoring? At best, you get inverter manufacturer software that shows pretty graphs but can't export data. At worst, you get screenshots of the inverter display sent via WhatsApp at noon every day.

• Payment Collection? Screenshots of bank transfers. If you're lucky, email confirmations. Reconciliation means scrolling through WhatsApp to find that screenshot from three months ago.

(And yes, I'm thankful my employer is way more advanced than this. But we're the exception, not the rule. And even we're just 30% of the way to where we need to be.)

The same week I watched this chaos unfold, I counted 47 new AI startups launched. Forty. Seven.

The Perfect Storm Just Arrived

Remember that fuel tax I mentioned? Every liter of petrol and diesel will soon cost 5% more. For businesses running on diesel generators 12 hours a day, that's not a tax—it's a solar sales pitch.

But NERC's proposed Net Billing Regulations is about to create a massive software opportunity nobody's talking about. Let me break it down:

What Net Billing Actually Means:

• Businesses can install 50kW to 5MW of solar

• Export excess power to the grid for CREDITS (not cash)

• Credits roll forward indefinitely

• Complex billing reconciliation every month

• Multi-party agreements (User, DisCo, NERC, NEMSA)

Now imagine tracking all this with Excel. Actually, don't imagine—that's literally what's about to happen unless someone builds proper software.

The China Lesson

That Fortune article about China's AI advantage? Everyone focused on the wrong part. They saw "China has excess electricity" and thought "infrastructure problem."

Wrong. It's a software problem.

China didn't just build more power plants. They built the digital infrastructure to manage, monitor, and monetize every electron. While the US is still debating grid modernization, China turned their entire energy system into a software platform.

Now look at Nigeria:

• 85 million people without grid access

• 40% of grid power lost to theft and inefficiency

• Zero unified platform for distributed energy

• Every solar company building their own half-broken monitoring system

• Payment collection that would make a 1990s banker cry

This isn't an infrastructure gap. It's a software gap. And it's worth trillions.

Shout-Out to Odyssey

Yes, Odyssey Energy exists. And yes, they're doing good work on project finance and deal origination. I've used their platform. It's solid.

But here's the thing: One platform isn't an ecosystem. It's a lonely pioneer in a desert that needs a thousand springs.

Odyssey handles maybe 1% of what's needed. Where's the:

• Operating system for mini-grids?

• Shopify for solar installers?

• Stripe for energy payments?

• Twilio for smart meter communications?

• Salesforce for C&I developers?

• QuickBooks for energy accounting?

• GitHub for energy system configurations?

Having one good platform is like having one good road in Lagos. Better than nothing? Sure. Sufficient for economic transformation? Not even close.

The Net Billing Software Gap

With NERC's new regulations, we need software that can handle:

For Prosumers:

• Track energy imported vs exported in real-time

• Calculate credits (at different tariffs!)

• Manage carried-forward credits indefinitely

• Handle premise transfers with credit preservation

• Interface with DisCos, NERC and all the potential 36 State ERCs, NEMSA

For DisCos:

• Manage thousands of net billing agreements

• Real-time monitoring of distributed generation

• Prevent the 30% network capacity limit breach

• Automated feasibility studies

• Credit reconciliation and escrow management

For Regulators:

• Real-time visibility into distributed generation

• Automated compliance monitoring

• Carbon credit tracking and allocation

• Performance analytics across the ecosystem

You know what we're going to use instead? Excel sheets emailed monthly. PDFs nobody can search. WhatsApp messages for coordination.

This is insane.

Why Energy Software Is the Real AI Play

Everyone wants to win at AI. Nobody realizes energy is the bottleneck.

Simple math:

• One GPT-4 query ≈ 0.004 kWh

• Nigeria's entire grid capacity ≈ 7,000 MW (on a good day)

• Actual available power ≈ 4,000 MW

• One modest AI data center ≈ 100 MW

You want to compete in AI? You need power. Reliable, scalable, software-managed power.

But the software stack for managing distributed energy in Africa doesn't exist. We're trying to build tomorrow's AI economy on yesterday's infrastructure managed by today's WhatsApp groups.

The Stack We should be Building

After three months of research (read: asking stupid questions to smart people), here's the software stack African energy actually needs:

Layer 1: Asset Management

• Real-time monitoring that works on 2G

• Predictive maintenance using local weather data

• Inventory tracking that doesn't require a PhD

• Performance analytics in actual Naira, not theoretical dollars

Layer 2: Customer Management

• KYC that works with Nigerian addresses

• Credit scoring for the unbanked

• Payment collection via USSD/mobile money

• Contract management that lawyers can actually use

Layer 3: Financial Infrastructure

• Project finance modeling for African realities

• Multi-currency settlement (try modeling a project with USD debt and Naira revenues)

• Automated FOREX hedging

• Carbon credit integration

• NET BILLING CREDIT MANAGEMENT (this is now critical!)

Layer 4: Grid Integration

• Load forecasting that accounts for "Nigerian time"

• Distributed energy resource management

• Peer-to-peer energy trading

• Grid stability services

• Anti-islanding protection monitoring

Layer 5: The Platform Layer

• APIs that actually work

• Data standards someone thought about

• Interoperability between systems

• An app store for energy

Total investment needed to build this? Maybe $50-100 million.

Total value created? Try $10 billion. Minimum.

The Unicorn Factory

Software businesses crushing soft costs in energy aren't just good businesses. They're generational businesses.

Look at what happened with solar software in the US:

• Lead generation platforms: Multiple unicorns

• Design software: Billion-dollar exits

• Financing platforms: IPOs

• O&M platforms: Roll-up targets at premium multiples

Now multiply that by:

• Larger addressable market (450 million people without reliable power)

• Faster growth rates (30-50% annually)

• Zero legacy competition

• Desperate need for solutions

• NEW REGULATORY FRAMEWORK REQUIRING SOFTWARE

Every single vertical in energy needs its Shopify, its Stripe, its Twilio. And nobody's building them.

Why Now Is THE Moment

Four things changed this year that make this inevitable:

1. The Economics Flipped Solar + storage in Nigeria is now cheaper than diesel generators. The new fuel tax just made it even more obvious. The market is about to explode from "nice to have" to "economically stupid not to have."

2. The Regulatory Framework Arrived Net billing regulations mean businesses NEED software to track credits, manage agreements, and prove compliance. This isn't optional anymore.

3. The Talent Exists Nigeria has world-class developers. I've seen the code. I've hired the teams. The same engineer building your AI chatbot could build the infrastructure platform that powers a continent.

4. The Capital Is Ready International climate funds are sitting on billions they can't deploy because there's no software infrastructure to track, monitor, and verify projects. Build the picks and shovels, and the gold miners will come running.

The Real Question

So why is everyone still building AI tools?

Simple: They're solving Silicon Valley's problems, not Lagos's problems. They're pattern-matching to Y Combinator, not Nigerian reality. They're building for the market they read about on TechCrunch, not the market they live in.

The biggest opportunities look like the worst ideas. Energy software in Africa looks hard, unsexy, and full of infrastructure headaches.

Perfect. That's exactly what a trillion-dollar opportunity looks like before it's obvious.

Your Move

If you're a developer reading this, you have two choices:

1. Keep building AI tools for a market dominated by OpenAI, Google, and Anthropic. Compete on features. Race to the bottom on pricing. Hope for an acqui-hire.

2. Build the software infrastructure for the largest energy transformation in human history. Own a category. Create generational wealth. Actually matter.

If you're an investor, the choice is even simpler: You can chase AI deals at 100x revenue multiples, or you can fund the picks-and-shovels for a gold rush that's already started.

And if you're like me – just someone trying to figure this out – start asking better questions:

• Why does every solar company build their own monitoring?

• Why can't I track my power consumption in real-time?

• Why does project finance still use Excel?

• Why isn't there a "NetBillingOS" for the new regulations?

• Who's going to build the credit management platform?

Because someone's going to answer these questions. And when they do, they won't build a unicorn.

They'll build a dozen.

The Uncomfortable Truth

Energy is how China won manufacturing. Energy is how the US won the internet. Energy is how Africa wins the next century.

But only if we build the software to unlock it.

Every day we waste building another AI wrapper is a day China extends their lead. Every brilliant developer optimizing ad-tech is a missed opportunity to power a continent. Every venture dollar chasing the current thing is capital that could have built the future.

I don't know who needs to hear this, but: Stop building toys. Start building infrastructure. The opportunity is bigger than you think, the timing is better than you know, and the impact is greater than you can imagine.

The gold rush is here. The regulations just arrived. The fuel tax just made it urgent.

Someone's going to sell the shovels.

Why not you?

—S

P.S. - If you're building energy software in Africa, I want to talk to you. If you're thinking about it, I want to convince you. If you think I'm wrong, I want to learn why. My DMs are open. If you're suffering with Excel-based project management, also email me. If you know React and want to help me not embarrass myself, definitely email me..

P.P.S. - Seriously, read those Net Billing Regulations. Schedules I through VII. That's not bureaucracy—that's a software requirements document disguised as policy. The opportunity is literally written into law.