Residency is seven days. Scale is earned.

In review

Most capital meets a company after it already exists. A deck. A round. A product that has survived long enough to be explained. That is a reasonable place to write a cheque. It is a late place to form the company.

Overture sits one step earlier. Overture Residency is seven intensive days for founder-ready people at the beginning of something — no idea, an early thesis, a prototype, or the first sign of traction. The question is narrow: should this founder and this problem become a company?

We call that Residency. Grow is still bounded: product, customers, distribution and the path to seed. Scale is when the company has repeatable traction and can compound. We will not scale the story ahead of the evidence.

Streams

We do not work on everything. The residency is pointed at four constraints that already sit in front of people in South Africa. Adjacent problems can enter if they are equally concrete. “Impact” is not a stream.

Work: unemployment remains the country’s hardest labour-market fact. The usual response is a programme. We are interested in companies that create work people will actually take — matching, tasking, quality, payment — at a cost the market can bear. Evidence is someone hired or paid who was not, on repeat. Not a pilot.

Transition: people need to move from one kind of work to another. The current path is slow, expensive and often dishonest after the certificate. AI should collapse the time and fiction between the person they are and the work they can do next. Evidence is different, better work — and someone who paid for the outcome.

Connectivity: bandwidth, devices and prepaid cost still decide who can participate. Products that assume fibre and a flagship phone fail where most people are. We want software that completes a real task on the network they already have.

Local markets: most small firms never become companies that can grow. The founder is the entire operating system. Credit and payments are already crowded. The opening is formation and operations — a local Atlas — so a real company exists in days, not months of offices.

AI

Tools collapsed the cost of making software. They did not collapse the cost of being wrong. A founder can now ship in a week what used to take a quarter. That makes a long, unfocused formation process harder to justify — and a short, honest residency more useful.

The systems that matter to us are no longer chat windows. They read a codebase, edit files, run tests, and hold a task long enough to finish it. That is why Residency can be seven days without being theatre. Execution accelerated. Signal still comes from a customer, a payment, a constraint in the real world.

If execution is cheap, selection becomes the work. We select for how people operate: agency, speed, a mind that can change, and the willingness to commit if they find something worth it.

Markets

South Africa is not short of problems. It is short of companies that turn those problems into something a buyer will pay for. Unemployment, expensive data, and small firms that never formalise are not themes. They are operating conditions. Formation that ignores them will produce demos. Formation that starts from them can produce companies.

Partners who sit close to those conditions — banks, telecommunications companies, enterprises, development institutions — can put problems, customers and distribution into the residency while the company is still being formed. That is a different moment from sponsoring a demo day.

Investors do not need to sit in every session. Overture Signal is a short recurring note on what emerges: product, customers, revenue, the change since last time, the constraint, the next milestone. Spend time when the evidence says it is worth it.

Developers

The stack is now chosen in week one, not in a later procurement cycle. That is why technology companies belong at formation. Cursor, OpenAI Codex and Vercel are used in Overture today. They are tools, not announced partners unless that relationship is confirmed.

A capable founder plus an agentic environment can produce a working product before the week is out. The constraint moves from typing to taste: what to build, who it is for, whether anyone comes back. That is a better constraint. It is also why we will keep measuring shipped product, users and revenue — not prompts.

Developers who can hold that loop — build, test, talk to a customer, change their mind — are the people the residency is for. They do not need a polished startup. They need to move when they find signal.

Outlook

The first residency is the proof, not a brand launch. Seven intensive days. Select founder-ready people. Expose them to problems and customers. Decide which founder–problem combinations should continue. Track what ships.

If that cycle produces signal, Overture grows. If it does not, we will know quickly. That is the same standard we apply to founders.