Why we're building six companies at once
Mind, learning, work, motion, climate and food. Six places where AI will either lift people up or leave them behind, and why SISU AI is building in all six.
People ask us one question more than any other: why six? Founders are told to choose one problem and go deep, and most of the time that’s good advice. We did choose. We chose people.
Picture a young woman in Tamale, in northern Ghana. In one ordinary week she might worry about a brother who has gone quiet, study for a certificate after work, wonder whether her job will look the same in five years and notice food prices climbing after a poor rainy season. Her life doesn’t come divided into industries, and AI is about to reach every part of it at once.
Every technology era is eventually judged by what it does for ordinary lives, and AI will be judged the same way. The test is simple: is a nurse, a student, a farmer or a shop owner better off because it exists?
So the question we ask is not which industry to enter. It is which human needs AI will reach, and whether it will lift people up or leave them behind. Six stand out: a healthy mind, the chance to learn, work with dignity, help in the physical world, a livable planet and enough safe food. SISU AI is building a company for each one.
Six companies is what focus looks like when you focus on people.
One platform, six frontiers
Building six companies would be reckless if each one had to start from scratch. None of them does. All six build on one shared SISU AI platform.
At the base is scalable inference infrastructure, which runs the models. Above it, a model aggregation layer brings frontier closed models and open-source models together, so no company is locked into one vendor. Agent harness frameworks, autonomous AI agents and the applications people use sit on top.
Alongside the platform is Alkebulan, our Pan-African large language model, multilingual by design and built to preserve what generic training data misses. That matters because close to half of the text crawled from the web is in English, and most African languages barely register.
A shared base means that what one company learns, all six inherit: models, tools, evaluation methods and hard lessons about building for low bandwidth and many languages. When one team works out how to make AI useful over a weak connection, the other five shouldn’t have to work it out again.
A healthy mind
In many countries, men’s suicide rates are three to four times higher than women’s. Behind that figure are fathers, sons, brothers and friends. Vibelane works on men’s mental health and workplace well-being, with a simple promise: Mind well. Live well.
AI4MentalHealth, its interactive platform, gives men resources to understand their mental health and thrive, with care that adapts to individual needs. Vibelane aims to bring that support to people in their native language and to remote communities. It is building a large language model focused solely on mental health, and it offers an enterprise solution for workplace well-being. If people are going to talk to AI about how they feel, it should understand how they speak and where they live.
The chance to learn
The World Economic Forum’s Future of Jobs Report 2025 expects 39% of workers’ existing skills to be transformed or outdated by 2030. When skills age that fast, learning can’t end at graduation, and a good mentor can’t be a privilege of the well connected.
Altera builds learning and mentorship pods that pair learners with industry leaders through real-world projects. Its four pathways are the Altera Fellowship, a cohort community; Altera Xperience, for one-on-one guidance; the Altera Accelerator, cohort support for startups; and Altera Academy, year-long, coached learning pods that use Teaching at the Right Level.
Altera’s AI models, co-developed with Kalon, are fine-tuned for personalized learning and include speech processing, so the AI speaks the learner’s language. A model can explain a concept at any hour, and a mentor can show you what a career looks like. Altera offers both. In 2025, its scholarship support came to more than $27,000.
Work with dignity
The same report expects 170 million jobs to be created and 92 million displaced by 2030, a net gain of 78 million jobs. Those numbers describe millions of working lives about to change. Dignity, in that shift, means people stay in charge of the tools that work beside them.
Rosanne AI builds for that future of work. Its agentic web browser, ROS, is made for Mac and iPhone and built on WebKit. It blocks ads and trackers and keeps your history encrypted on your device, with no account needed. Agents are coming to ROS.
Rosanne Agentic OS lets you talk to agents with clear roles that plan, act and report back, with voice AI built in. You set the permissions and approve the key steps. The whole flow fits in four words: Ask, Plan, Approve, Report. The agent does the work, and you keep the judgment.
Help in the physical world
Much of what keeps people alive happens away from screens. Crops have to be harvested, goods carried and people cared for by hands that can lift, reach and move. Software on its own can do none of that.
Pibotix is our company for physical AI and robotics. Its motto is Perceive. Reason. Act. It traces an arc from symbolic AI to perception, then generative and agentic AI, and now toward physical AI: intelligence embodied in action. As Pibotix puts it, intelligence must move.
Pibotix builds in the open, with a public repository on SISU AI’s GitHub. When robots arrive where people live and work, those who need that help most should be among the ones who build it.
A livable planet
Africa accounts for less than 4% of global greenhouse-gas emissions, yet it faces some of the most severe climate impacts. The people who did little to cause the problem deserve the best tools to face it.
Vaxus is where AI meets Earth, with a focus on carbon accounting and climate awareness. Its carbon platform promises “one platform, all of carbon,” and it also offers consultancy. VAXUS GPT, launched in 2023, builds that awareness in local languages. With USSD and SMS support, Vaxus works online or offline.
It is also working to open access to carbon markets for Africa’s ecosystems, including a market for more than 12,100 hectares of forest. In the age of AI, climate knowledge shouldn’t stop at the edge of a data plan or a dominant language.
Enough safe food
According to the UN’s Food and Agriculture Organization, about 868 million people in Africa were moderately or severely food insecure, 342 million of them severely. Having enough food is half of the need. The other half is knowing it is safe to eat.
Dinera uses AI for food safety, food security and sustainable packaging, from farm to fork. Its tools include Dinera Chat, a QR platform, and e-learning and research platforms. Its app is built to work online and offline, and people can also reach Dinera through USSD.
Food that is safe to eat should also come in packaging the planet can live with. And what science knows about keeping food safe and plentiful should reach every link in the chain, including the least connected.
Built for the next billion
Look across the six companies and the same design choices keep appearing. Vaxus supports USSD and SMS, Dinera’s app is built to work offline, Vibelane aims to reach people in their native language, and Altera’s principles include being edge aligned. Those choices follow from how people actually live. Many people live within reach of a mobile broadband signal but don’t use mobile internet, often because of the cost of data and devices.
Responsible AI is part of the same discipline. Research has shown that safety guardrails tuned in English can be bypassed in low-resource languages, the ones with little text online. Alkebulan’s v0.0.17 release, “Accra,” includes responsible-AI guardrails tuned with community reviewers to flag cultural misinterpretations.
Our companies, and our partner OpenAI4Africa, work across Academia, Government and Industry. We call it the AGI approach. And we build in the open, with a public, MIT-licensed repository for each of the six companies at github.com/SISU-AI. Our developer page is the place to start.
This matters most in emerging markets, where most of the next billion users live. Africa alone shows the scale.
We’ve seen what happens when technology fits the way people live. M-Pesa launched in Kenya in 2007 and helped make mobile money mainstream across the continent. AI will shape the next century, and SISU AI exists so emerging markets can be architects of that shift.
So yes, we chose. We chose people. And people’s needs don’t take turns.