In February 2026, mLab Southern Africa published a small piece of news with a large implication. Two CodeTribe graduates had joined The Startup Legal. Karabelo Nthoroane became a frontend lead; Lebogang Mylas joined as a backend support engineer. Mylas described his starting point with refreshing plainness: “Before CodeTribe, I couldn’t confidently use a laptop.” His final academy project helped him earn the position. Somewhere between those two facts lies the reason mLab exists.
- For learners: practical coding, cloud skills and preparation for work.
- For founders: help with customers, business models and early products.
- For partners: a local organisation that turns programme funding into delivery.
The organisation began in 2011, when a consortium including World Bank infoDev, South Africa’s science department, CSIR and The Innovation Hub helped establish a mobile applications laboratory. The proposition was understandable: support the people building the mobile economy. Over time, the question became more demanding. Who would build the products? Who would buy them? And what happened to the enthusiastic young person after the event photographer went home?
The applause had a short shelf life
A 2018 World Bank account describes the adjustment. Early hackathons and pitch competitions brought attention and confidence, but then CEO Derrick Kotze warned that those events did not necessarily deliver tangible returns. mLab became more selective about teams and more deliberate about progress. It also separated two needs: inexperienced young developers needed practical skills; entrepreneurs with industry knowledge needed help making viable solutions.
“You need talent. You need time. You need skills and processes.”Derrick Kotze, then mLab CEO, 2018
The lesson is useful well beyond South Africa. An event can introduce people. Sustained support must equip them to do something together. mLab’s response included connecting young developers with experienced entrepreneurs and helping build solutions around identified customer needs. That creates a rather different job for an accelerator: less judging from the balcony, more working at the bench.
A portfolio with fingerprints on it
CodeTribe, launched in 2014, gives that idea a practical form. Its skills programme covers Android, iOS and web development, cloud technologies and scrum agile project management. Participants develop a portfolio and gain work experience. The organisation says it aligns the curriculum with business partners’ needs. For a learner, the appeal is evidence of work that another person can inspect.
Nthoroane’s graduate account makes the daily mechanics visible: Git and GitHub, merge conflicts, teamwork and presenting projects. These are modest-looking skills until a deadline arrives. A developer who can explain a change and work within someone else’s codebase is easier to employ than one whose expertise never leaves a tutorial.
The academy also has a geographic purpose. Its 2016 expansion with The Innovation Hub took CodeTribe into Soweto’s eKasiLabs centre after an initial 36-student validation phase. That intake offered 60 full bursaries across Soweto and Tshwane. The financial barrier was addressed through sponsorship; the work still demanded time and commitment. Today’s general skills criteria describe programmes lasting six to twelve months and require applicants to live near an academy.
Give the business a young technologist
STEP UP offers another way across the distance between training and income. Its first cohort, launched with Edunova in February 2025, involved 17 young people from Daveyton. They trained for six months as system administrators, then entered twelve-month placements with small and medium-sized enterprises. The business gets digital capability; the trainee gets a workplace where capability has consequences.
In September 2025, mLab announced GIZ support for STEP UP II, targeting 100 young people across Gauteng, Northern Cape, KwaZulu-Natal and Limpopo. That was a programme target. The design added employer-aligned cloud certification, mentorship, workplace readiness and digital marketing. It also anticipated freelance work. A useful training programme must understand that the next opportunity may come from a small business rather than a corporate graduate scheme.
Rural and township access appears elsewhere, too. A renewed July 2025 agreement with MICT SETA followed four introductory Internet of Things cohorts: three in the Northern Cape and one in Richmond, KwaZulu-Natal. Location is part of the delivery problem, alongside the syllabus.
First find someone who needs it
For founders, mLab offers gap analysis, go-to-market planning, mentorship, MVP development and programme-specific seed grants. Its general enterprise guidelines favour young businesses with evidence of market validation. These requirements make sense: a working product can still be an expensive answer to a question nobody asked.
The 2026 mLab Digital Enterprise Accelerator makes the sequence explicit. Announced in June for July 2026 through March 2027, it starts with eight modules over sixteen weeks on the Conecta platform. Participants examine customers, value propositions and business models. Twenty enterprises are then selected for further support. A final pitch event is scheduled for March 2027, with the top three sharing grant funding.
mLab also develops digital solutions with partners. Its published examples include an agricultural knowledge platform, the youth opportunity app mPowa and miniSASS, which assesses river health through the small animals living in the water. The subject matter is pleasingly unglamorous. A river insect has little interest in a pitch deck, but it can supply useful information.
Who carries the bill?
mLab is a nonprofit company with public benefit organisation status and Level 1 B-BBEE credentials. Its market position combines skills delivery, enterprise development and technology services for institutional partners. Learners and founders receive support; public bodies and corporate partners help make programmes possible. The historical World Bank account says supported startups surrendered neither equity nor royalties.
The distinction matters when reading its numbers. mLab’s undated impact page reports 500 developers trained and 200 startups supported. It attributes R150 million-plus in fundraising to supported startups. That is money raised by beneficiaries, rather than the organisation’s own revenue. The useful takeaway is the linked model: train talent, test demand, build something inspectable and connect it to work.
Applicants should read each current call carefully. Prior qualifications, proximity and availability affect access; startup selection is competitive. An employer needs the capacity to supervise a placement. A founder still needs customers. mLab’s offer is practical help with those journeys. The graduate project that earns an interview is a good place to see the idea working.