It still feels strange to call it a company, but it is one. I started Aveling Research, a quantitative research firm founded in Cameroon.
My background is in software engineering and AI engineering. Aveling grew out of applying that background to systematic trading research. I began testing trading ideas, building the data pipelines needed to evaluate them, and recording what survived realistic costs and out-of-sample testing.
Most of the ideas failed. Some looked promising before transaction costs. Others worked in one period and disappeared in the next. One study of dividend capture on the West African regional stock exchange, the BRVM, passed the full paper-validation process, although it still requires confirmation from a broker and a live micro-test before it can become operational. The failures mattered because they showed that the research process could reject attractive stories instead of finding a way to justify them.
That work also clarified where Aveling should begin. The largest quantitative firms compete in markets where speed, colocation, capital, proprietary data, and execution infrastructure matter. They have spent decades building those advantages. Starting by competing against them on the same terms would make little sense.
African economies and markets present a different research surface. They produce financial, economic, environmental, and operational data, but much of it is slow, fragmented, and difficult to extract. The information may exist in a report or an irregular publication rather than a clean feed that can be loaded directly into a model. This makes it less attractive to firms built around liquid markets and fast signals.
Aveling Research is being built around that gap. The advantage we’re looking for isn’t faster execution. It is the ability to collect information that is difficult to use, preserve it as structured data, and test whether it contains a relationship that survives outside the original sample.
The engineering has a specific role in this process. Software collectors preserve the original source and turn it into reproducible datasets. AI helps extract and classify information from documents that were not published for machines. Statistical tests measure whether one series predicts another, while cost models determine whether a relationship could be traded. Humans still decide which hypotheses to pursue and whether any result should carry capital.
The company begins as a research firm, not a fund. Its first job is to build the data library, the research infrastructure, and a record of hypotheses that passed or failed under the same standards. Some datasets and models may eventually be useful to other firms. A relationship that survives validation may become one of our own systematic trading strategies. Until either happens, it remains research.
I started the company, and we are currently a team of three. We will work mostly remotely, and the three of us have the capacity to build the research system and carry the work through. We are building a firm where engineers and quantitative researchers can build collectors, audit datasets, model relationships, and reject their own results when the evidence doesn’t hold.
The initial focus is Africa, but the company isn’t named after one geography or one strategy. The research can extend into other markets where the same methods provide an advantage. Competing in the large global markets remains possible, but it isn’t the starting point.
For now, the work is direct: collect difficult African data, make it usable, test whether it predicts anything, and keep only what survives.
If this is the kind of work you want to help build, email me.