There is a moment that Mutembei Kariuki most likely finds difficult to forget. A barbershop owner in Nairobi sent him a WhatsApp message containing a stack of M-PESA transaction statements at some point in 2020, during a pandemic that had closed stores and frozen supply chains. It was more than just a spreadsheet. The shop owner had never once counted, tracked, or acknowledged the 180 devoted customers, which was a revelation. A barbershop that had no idea who its regulars were. Even though it seems insignificant, that detail reveals nearly everything about the gap Fastagger is attempting to bridge.
Founded in 2019 by Mutembei Kariuki, Jude Mwenda, and Stephanie Njerenga, the Nairobi-based startup has been doing something that larger, better-funded companies have found difficult to accomplish throughout the continent: enabling artificial intelligence to function for regular African businesses on regular African devices with regular African internet—that is, sometimes no internet at all. It sounds humble. It isn’t.
| Category | Details |
|---|---|
| Company Name | Fastagger |
| Founded | 2019 |
| Headquarters | Nairobi, Kenya |
| Co-Founders | Mutembei Kariuki (CEO), Jude Mwenda, Stephanie Njerenga |
| Core Product | Auni — AI-powered business intelligence assistant |
| Key Integration | M-PESA (Safaricom) |
| Official Launch of Auni | March 2024 |
| First Month Adoption | 600+ MSMEs |
| Accelerator | Google for Startups Accelerator: AI First (2023) |
| Target Market | MSMEs across Sub-Saharan Africa |
| Reference Website | blog.google/en-africa |
The MSME landscape in Africa is vast, somewhat disorganized, and severely underserved by the enterprise software that Silicon Valley usually produces. These are the market vendors, roadside pharmacies, and small logistics companies that use M-PESA to process thousands of shillings every day with little to no insight into what those transactions are telling them.
This is where Google has its infrastructure. Here are users of Meta. However, neither has solved the issue of making AI actually helpful to a Kisumu shopkeeper whose phone is a 2019 Android with 2GB of RAM and a patchy data connection.
The app Auni is Fastagger’s response. After two years of development and testing, Auni was formally launched in March 2024. It pulls financial statements, integrates directly with M-PESA, and transforms them into information that a real business owner can use, such as customer trends, revenue patterns, and loyalty signals. Over 600 MSMEs used the platform in just the first month. That’s not viral growth in the Silicon Valley sense, but it’s significant traction in a market where onboarding is a real challenge and trust must be earned.
The team’s journey there adds intrigue to the narrative. Kariuki and his co-founders were labeling data for South Korean and American businesses prior to Auni, the Safaricom partnership, and all of this. The idea of African engineers training AI systems for overseas markets and then using that same expertise to create something completely for home has a subtle poetic quality. There was always competence. The course shifted.
In 2023, Fastagger joined Google’s own accelerator program, which, depending on your point of view, is either ironic or just practical. Using the cloud credits, they created foundational document models especially for African financial documents. These models were created from scratch to reflect the actual appearance of M-PESA statements rather than being modified from Western templates. It may not seem important, but that distinction is crucial. The majority of AI tools used in Africa are modified versions of those used elsewhere. This is where it all began.
The precise number of markets in which Fastagger is operating at full capacity is still unknown, and the business hasn’t made a big deal out of its growth. However, their architecture, which is low-bandwidth, offline-capable, and integrated with the mobile money rails that currently facilitate daily commerce throughout the continent, is better suited for replication than cloud-dependent tools. Running machine learning models directly on a basic smartphone is not a workaround in markets where data is expensive and electricity is unreliable. It’s the entire approach.
Observing this from the outside, it seems that Fastagger embodies something that the African tech sector has been subtly working toward for years: startups that design from local reality outward rather than adapting foreign products for local use. Although it sounds philosophical, adoption rates reflect the difference. It manifests as a barbershop owner who at last understands who his clients are.
There are genuine concerns about whether Fastagger can maintain this momentum, raise the necessary funds without losing its operational focus, and expand into twelve or more markets without the unit economics collapsing. The startup graveyard in Africa is quite large. Companies that appeared stable have failed due to currency volatility alone. However, it is more difficult to replicate the foundation that Fastagger has established, beginning with mobile money data and true infrastructure thinking.
Big Tech has the means. The insight is in Fastagger. That is, in fact, the more resilient advantage in sufficient markets.