Opinions on AI are rightly divided. Builders and innovators who have championed AI’s growth see it as the future – the future of work and the future of saying “no” to mundane, repetitive tasks. On the other hand, users are divided into those who support AI adoption and those who worry about the impact of the cutting-edge technology on jobs.
But this hasn’t slowed down AI’s rapid progress, which has been anything other than accidental futurism.
In Africa this innovation wave is rapidly gaining ground thanks to ChatGPT, which has sparked startup concepts that are on the edge of replacing jobs and removing mundane tasks from businesses.
The rise in AI-powered “Swiss Army’ tools
Recently, some of these startups have appeared in Nigeria: a0dev, a Nigerian startup that joinedY Combinatorin 2025 allows you to build a React mobile application from scratch by using prompts. And now, Egypt’s Stakpak, – through the use AI agents – helps software and DevOps engineering set up production-ready Infrastructure much faster.
At least 45% developers use six core products to ship their product, but configuring them can be complex, George Fahmy told TechCabal. Each tool has its own language. Teams must read hundreds of pages of documentation to get a product out to customers. Only about 3% have the skills needed to manage DevOps full-time.
There is a global shortage of DevOps Engineersmaking the few professionals who are available high-maintenance. According to Payscale, an online salary aggregator for Nigerian engineers, the average engineer earns N250,000 ($163) per month. This overhead cost is on top of the software engineering teams companies must hire.
In a company with a heavy infrastructure, such as eTranzact,entry-level DevOps Engineers earn between N250,000 and N400,000 ($163 to $260) per month, while mid-level professionals make over N600,000. ($390), depending upon performance and negotiation.
Some firms hire cloud engineering consulting firms in order to offset the costs of paying full-time salary. Clutch states that these firms charge between $10k and $50k per project.
With platforms like Stakpak and a0dev on the rise, founders can become the startup equivalent to a Swiss Army Knife, shipping products quicker, hiring fewer talent and saving money on overhead costs. AI development tools make it easier to turn ideas into products, but they may also force talent to evolve and pivot.
Scaling downside, building upside
Due the high cost to retain DevOps talents, some startups train software engineers to take on hybrid roles. This can work in the short-term, especially since startups are shipping fewer features, and they often use serverless technologies, reducing their need for infrastructure management.
As workloads increase, engineers need more control of infrastructure. DevOps is required to transition from serverless architectures to managed Kubernetes, or other scalable architectures. If they can get past this hurdle, the growing infrastructure demand will disrupt the workflows of engineers who split their time between DevOps and development, slowing product iteration, and increasing time to market.
Some startups are turning to AI powered DevOps Tools to ease this burden. These tools automate infrastructure setup, management and maintenance, reducing the requirement for DevOps engineers at the early stages.
Stakpak, for example, has received early traction in Egypt from early-stage startups such as Credify Paymob and Zammit. Its AI agentic tool costs only $50 per month.
Stakpak is a great tool for early-stage engineering teams. However, its application in large enterprises is still a bit uncertain. Stakpak’s artificial intelligence is used in critical DevOps tasks, so reliability is essential. Mistakes with infrastructure setup can lead to security risks, cloud costs, or downtime.
Aaron Adetunmbi is a Nigerian software programmer. He said that it might take longer for larger companies to adopt agents. “They have complex processes and the technology is so new that they don’t know where to start.” LLMs are prone to ‘hallucination’, which can lead them to make costly mistakes.
Fahmy says that Stakpak is built with a mixture of closed-source LLMs and open-source LLMs. However, the startup has its own benchmarks for evaluating its data, ensuring its AI agents’ reliability. Fahmy claims that manual evaluation checks can cost up to $400 per experiment. This makes AI-powered platforms costly to build.
He admits it’s been hard to sell to large enterprises. The startup is now focused on early-stage companies, with the potential to add larger clients in the future. Fahmy said that Stakpak, which was launched in 2024, has already begun to generate revenue. Fahmy said that “our tool is niched and we’re still validating our best clients.”
“We haven’t reached product-market fit but we are iterating our products and generating revenue from clients who have been paying for over a month and a half after we released our first version.” He believes that the adoption of agentic deployments tools like Stakpak in Africa may take some time.
Is the future of jobs in danger?
Based on a survey of more than 30,000 software developersa 2024 Stack Overflow Reportfound that the majority of them expect AI to impact their workflows by 2025. Around 79% of software developers believe AI will have a major impact on how they document, test, and release code.
Meanwhile, the term “vibe coding” has gained traction, referring to entrepreneurs rapidly building and deploying software-as-a-service (SaaS) apps using prompts. This trend is extending beyond software development into areas such as marketing, DevOps and design, demonstrating AI’s increasing versatility.
Beyond the startups’ utility, there is investor confidence in these startups. Stakpak raised $500,000 from P1 Ventures with investors Digital Currency Group and 500 Sanabil. Angel investors also participated.
Adilege, and raiseda undisclosed sum from a group YC alumni through Pioneer Fund. Evans Akanno’s BootstrappedVzy has not raised VC dollars but has received more than $50,000 in funding from friends and startup accelerators.
In Africa, more founders build AI agentic tools that solve industry-specific issues in logistics, finance and customer service. This affects more talent.
These agentic development tools, despite the fear that AI agents will replace jobs, are not intuitive. These tools are not intuitive for users or entrepreneurs with no coding experience. This makes it important to hire developers who have coding skills. There is a learning curve, however, for traditional engineers. They must become more comfortable using AI agents to develop and deploy their processes.
According to AkinkunmiTokede, DevOps Engineer, “AI agents will be the next big thing in tech. 2025 will be their breakthrough year.” “AI-powered assistants are going to make our lives easier in tech.” AI is often cited as a way to replace engineers, but that’s not what I see happening, at least for professionals with the right skills. AI agents are not here to replace engineers, but to assist them. They save time and money for businesses and engineers. The rise of AI agents shows a shift in the way software is developed, but also in who is building it and how. While the fear of job displacement continues, the reality is nuanced. AI is not eliminating developers. It’s redefining roles. For those who are willing to embrace the change, the future doesn’t mean being replaced. It’s about staying in front.
