Zambian AI communications startup, Caantin, wants to reduce the cost of making phone calls for businesses by using AI voice agents that can have these conversations at scale.
Communication is the lifeblood of many enterprises. From fintechs nudging users to complete signups or pay back loans to FMCG brands chasing daily orders from mom-and-pop shops, phone calls are a critical part of daily operations. Traditionally, that job has fallen to large or outsourced call centers, or in-house support teams that come with overhead costs, hiring headaches, and natural limits—there’s only so many calls a human can make in a day.
Caantin, a new AI communications startup, thinks it has a better way: AI voice agents that can have these conversations at scale.
Launched in 2021—Caantin’s first iteration was Topup Mama, a procurement and management software for restaurants—and self-described as an AI communications company, the Zambian startup says it wants to reduce the cost of making phone calls for businesses.
“Businesses struggle to scale for a number of reasons, and when you peel into it, you end up with a communications issue—the cost of communicating,” said Njavwa Mutambo, CEO and Co-founder of Caantin. “We help businesses communicate with customers in a way that is intelligent, contextual, and cost-effective.”
had a go at using Caantin’s incredibly human-sounding AI voice bots by acting as a mom-and-pop shop trying to place orders. Save the few gaps in communication, the conversation felt really natural. The call included the intonations, pauses, and inadvertent interruptions of a real life conversation. The AI chatbot allows you to have conversations accented in major African languages—Igbo, Hausa, Swahili, and Yoruba.
Mutambo claims the startup has helped the businesses it serves increase their efficiency. For example, Mutambo says Caantin helped Nigerian fintech Cowrywise achieve 100,000 calls with one employee, a feat that would normally be achieved with 30 employees in a 3-month period.
Voice is the entry point
While AI use cases on the continent are still nascent—largely due to the high cost of building and running models—Mutambo believes voice AI will be the most dominant use case of AI in Africa. With low smartphone penetration, inconsistent internet access, and high illiteracy rates in many regions, Mutambo sees voice—not text or chat—as the most practical interface for AI adoption at scale.
“Voice is how AI will be distributed in Africa,” Mutambo said.
But beyond just building voice bots, he argues the key is developing context-specific use cases that align with how enterprises operate on the continent. For Caantin, that means designing voice agents that can drive collections, complete sales flows, and reduce the cost of customer support—at scale.
For example, Caantin’s voice bot can prompt payment during calls using webhooks for integration with payment providers like Paystack and Flutterwave.
Caantin is also looking to add a feature that would analyse tons of phone conversations and deduce insights that can be used for business decisions.
How Caantin makes money
Caantin generates revenue by charging per second for phone calls, operating much like a telecom provider. In South Africa, it charges four rands (2 cents) per second; in Nigeria, about ₦117 (7 cents) per minute—roughly nine times the rate of a local telco.
Mutambo defends the pricing by pointing to the broader costs businesses typically incur to run large customer support operations.
“When you’re doing the math, you have to factor in staff salaries, power, internet—all the overhead it takes to run a call center,” he said. “Most chief commercial officers we work with see AI as a cost-cutting tool, not a cost center.”
In Caantin’s pitch, the higher per-minute price is offset by the scale, consistency, and 24/7 availability that AI voice agents offer—without the hidden costs of managing human teams. The company claims it is currently profitable.
Competition and differentiation
Caantin’s main competitors are traditional call centers—Business Process Outsourcing (BPO) companies—and global players like YC-backed Bland AI. Unlike its global competitors, Caantin is built with an African context. The chatbot understands local dialects, accents, and speech patterns more effectively than generic language models.
Caantin’s plan to offer an analytics feature puts it in direct competition with ToumAI, a Moroccan AI startup that works with businesses, including telecom providers, banks, and call centers, to gather and analyze voice data.
Fundraising and future plans
The startup raised an undisclosed round from Ventures Platform earlier this year. While Caantin’s goal is to make startup customer support more efficient, we asked if that spells the end for BPOs.
“BPOs will need to adapt or struggle to survive,” said Mutambo. Caantin’s AI isn’t perfect yet—but when it is, call centers won’t just be in trouble. They’ll be lunch.
