Applied AI Intern

Cape Town Based | Scale-Up | 3 month contract | Jem X | Hiring Manager: Rob Gardner Jem is on the hunt for Applied AI Interns based in Cape Town. Most graduate roles put you in a queue. You shadow someone for six months, you get handed the work nobody senior wants, and you learn how a big company protects itself from risk. This is the opposite of that. We are building what we believe will be the best applied AI team in South Africa, and we are hiring a number of people into it. You will not be shadowing anyone. In your first week, you will ship something that real people use. You will sit metres from the founders. You will work on problems where nobody in the building knows the answer yet, with tooling that did not exist eighteen months ago, at a company that has decided to bet itself on AI. If you are the kind of person who has been building things since before anyone paid you to, this is the role you have been waiting for. (Subscribe here to our CEO and founder's Substack, where he is documenting out journey towards becoming an AI-native company.) 💡 Who we are Jem is a Series A scale-up building WhatsApp-native HR, payroll and financial services for South Africa's frontline workforce - the people who clean, guard, cook, drive and serve. Our customers include McDonald's, KFC, Bidvest and Servest. When we ship something well, hundreds of thousands of people who have historically been last in line for good software get it first. Inside Jem sits our AI team, split into two halves. You will join one of them. JemX is our external-facing AI services and consulting division. We build bespoke AI solutions for some of the largest employers in the country - agents, workflows, integrations, dashboards, whatever the problem actually needs. This work is inbound and growing without us chasing it, and it puts you in rooms with executives at serious companies. Jem Labs is our internal AI team. Labs engineers embed directly into Jem's machines - Sales, Customer Success, Product etc - and rebuild how the company works from the inside using AI. This is not internal tooling as a chore. It is the thesis that every team in a modern company should have a technical person inside it, and you would be one of the first people in the country hired explicitly to prove that. Both halves are led by Rob Gardner . Both sit in the same room. You will be told which half you are joining, and you will see the other half's work every day. 💡 What you will actually work on Depending on which half you join: Applied AI builds for major enterprise clients: agents, workflows, integrations, data pipelines Embedding into a Jem machine and rebuilding how it operates - taking work that consumes a team's week and making it take an afternoon Jem Brain, our internal intelligence layer, and Jem Mini, our new payroll product Rapid prototypes that decide whether a product idea is worth committing to Clean, usable frontends for the tools and demos you build Because the team is small, your work will span AI/ML engineering, general product building and a bit of design. Whatever it takes to get something real in front of a real user. 💡 Who we are looking for We are going to market-wide and hiring narrow. The bar is very high, and we would rather leave a seat empty than fill it with someone who is merely good. We want the best . Real depth in AI and ML fundamentals. You understand what is happening inside the models, not just how to call them - training and fine-tuning, evaluation, why a model fails and how you would know. A degree in CS, AI/ML or a related field is one way to show this, but we care more about the depth than where it came from. chnical. We need actual fundamentals, not prompt engineering Highly proficient with Claude Code and coding agents. Fast, confident, and opinionated about how to use them Evidence you build things nobody asked you to build. Side projects, shipped work, a portfolio, a repo you are proud of. The strongest signal is a hard project you chased past the point where anyone was making you A first-principles, mathematical instinct. T he person who spots the constraint nobody else saw The ability to explain a hard technical decision simply. You will be in front of clients and non-technical colleagues High energy, genuinely curious, and comfortable in ambiguity 💡 Our process Deliberately hard, because the bar is the whole point. We open applications wide and expect several hundred. We interview deeply - expect to be asked to walk through the hardest thing you have ever built and then pushed two layers past where most interviews stop. Shortlisted candidates will come into the office for a full assessment day, where you will spend the day with our team solving a genuinely difficult, genuinely open problem alongside the other finalists. You will find it demanding. If it is your kind of thing, you will also find it the most fun you have had in an interview process. 💡 What You Can Expect From Us We offer a hybrid, flexible work environment, but we expect you to be in the Cape Town office, at least 4 days a week (12 Dreyer Street, Claremont) R30,000 stipend Lunch is provided 4 days a week when you're in the office, and you will have access to an EAP programme during the three-month contract and a MacBook Pro to work on. We respond to all applications. You can expect feedback on your application, even if we decide not to move forward. Do well, and there is a permanent role (and a ridiculous growth opportunity) waiting at the end of the contract - along with a real claim to being one of the best young applied AI engineers in the country. 💡 Why this is the opportunity Three months here will teach you more than two years almost anywhere else - and we mean that as a measurable claim, not a slogan. Real enterprise clients. Frontier tooling. Direct access to the founders. Colleagues who will push you and expect to be pushed back. The market is only now waking up to how scarce genuinely AI-native engineers are. Within a couple of years, everyone will be hunting for these people. The engineers who are exceptional by then will be the ones who spent this window building relentlessly, in the open, on hard problems, next to other people doing the same. That is what we are building here, and this is your chance to be in the first cohort of it. If that sounds like you, we're so excited to take you through the interview process.