The Founder Before the Startup

Joshua Barimon is a serial builder who started his first company at 16 and has experimented across several technology businesses, including social media management, web hosting, graphics design, and an African SME-focused fintech venture among others.

KrosAI emerged from a much more specific infrastructure problem.

While working on localised voice models at a previous company, Barimon’s team successfully built an AI voice agent but discovered that connecting it to a Nigerian telephone number was far harder than building the AI itself.

That experience eventually became the foundation for KrosAI.

---

The Core Problem

The rapid development of voice AI has created an interesting imbalance.

Platforms such as ElevenLabs, Vapi, and other voice-agent infrastructure providers make it increasingly straightforward to build the intelligence layer.

But actually deploying those agents into local telephone networks across emerging markets can involve telecom providers, KYC requirements, local regulations, number provisioning, and technical differences between carriers.

The founder experienced this first-hand.

His team spent roughly four months getting its first local phone number, followed by another month figuring out how to connect that number to the AI agent, because different telecom providers handle audio transmission differently.

The implication is important. The bottleneck is no longer necessarily the intelligence. It is deployment.

A voice AI agent that cannot reliably reach the person it is supposed to call has limited practical value.

---

The Strategic Decision Layer

KrosAI’s strategic decision was not to compete directly with the companies building the underlying voice AI models.

Instead, it sits between the AI agent and the telecommunications infrastructure.

The platform connects to different voice AI providers while handling the local infrastructure required to make those agents operational in emerging markets.

In practical terms, KrosAI wants developers to build their agent wherever they prefer, then use KrosAI as the deployment and connectivity layer.

The phone number is therefore only the entry point.

KrosAI’s broader ambition is to become the deployment layer for voice AI across emerging markets, with a promise of getting users from KYC to a live local number in under five minutes.

That is a fundamentally different proposition from simply providing another voice-agent builder.

An organisation conducting a survey across Kenya, Uganda, Rwanda, and Tanzania could deploy voice agents to conduct thousands of calls.

KrosAI’s role is to allow those agents to call people using local numbers, rather than appearing as unfamiliar international numbers that recipients may be less likely to answer.

The same infrastructure can potentially support customer service, banking, government services, NGO programmes, and other high-volume workflows.

---

The Ecosystem Context

KrosAI is entering an ecosystem where the AI layer is developing extremely quickly, while the physical and regulatory infrastructure underneath it remains fragmented.

Africa alone presents dozens of national telecom markets, different regulatory regimes, and different carrier requirements.

The founder’s argument is that scaling across these markets cannot simply be treated as one technical deployment problem replicated fifty times.

Each market introduces its own relationships, KYC processes, regulations, and telecom infrastructure.

This fragmentation helps explain why large global providers may not automatically solve the problem. It is technically possible to provide voice AI globally. It is considerably harder to build the local relationships and regulatory infrastructure necessary to make it reliable across emerging markets.

KrosAI is effectively betting that this complexity itself can become a competitive moat.

---

Observed Signals

The problem emerged from actual product use.

KrosAI did not begin with a broad thesis about voice AI infrastructure. It emerged after the founder’s team built a working voice agent and discovered that connecting it to the local telecom network was unexpectedly difficult.

That gave the company a concrete infrastructure problem to solve rather than an abstract market thesis.

Early commercial traction is meaningful.

At the time of the interview, KrosAI had over 60 paying customers across its markets, with enterprise and NGO use cases involving large-scale automated calling already active among them.

Security is becoming part of the product.

KrosAI is moving beyond connectivity. The company was developing real-time fraud detection designed to identify suspicious patterns during AI calls and alert users or potentially terminate calls through its dashboard, API, or MCP integration.

That potentially moves KrosAI from being simply a connectivity layer toward becoming part of the trust and safety infrastructure surrounding voice AI.

---

Open Variables

The opportunity is large, but so is the operational complexity.

Its biggest constraint is currently regulation and market access. The founder described backlogs of customer requests for markets including Malawi, Madagascar, Egypt, and the UAE, because entering each market requires groundwork and trust-building.

On the other hand, the founder is explicit that phone numbers are only the beginning. The longer-term ambition is to become the default deployment layer for voice AI across emerging markets.

The founder also identifies distribution as a lesson from building KrosAI. Around 90% of users had reportedly found the platform organically through content and word of mouth, while he believes he should have approached larger potential customers much earlier.

---

Why This Matters

KrosAI highlights a broader issue in emerging-market AI. Access to powerful models does not automatically translate into usable AI infrastructure.

If AI agents are going to interact with people who may have limited internet access but do have mobile phones, voice becomes an important interface.

But making that interface work reliably requires telecommunications access, regulatory compliance, identity controls, and local market relationships.

That makes KrosAI potentially interesting not simply as a voice-AI company, but as an infrastructure and market-access layer for AI adoption.

For investors and DFIs in particular, the question is whether KrosAI can turn today’s fragmentation into a scalable network.

The more markets it connects, the harder it becomes for a new entrant to reproduce the relationships, integrations, and regulatory knowledge required to operate across them.

---

Final Strategic Takeaway

KrosAI is built around a relatively simple observation. The future of voice AI will not be determined only by how intelligent the agents become, but by where those agents can actually operate.

The company’s bet is that as voice AI becomes commoditised, the harder problem will move downstream to connectivity, trust, regulation, and local deployment.

If KrosAI can make that complexity disappear for developers and enterprises, its most valuable product may not be the phone number at all.

It may be the infrastructure that makes voice AI usable across markets that global AI platforms were never designed to serve.

---

This article is drawn from an in-depth founder interview conducted by Afriq IQ with Joshua Barimon, CEO & Founder of KrosAI. Selected insights and observations are published here.