Beyond The Annotation Box: A Woman's Journey Through AI in Africa

By Sharon Iphy

A journey of a 1000 miles begins with a first step. I did not just stumble into AI - I built my way into it, label by label, task by task. I’m a Nigerian woman in Ghana, and I have spent the last four years growing from a Data Labeler to a Product Associate, shaping the way machines think and humans collaborate. This is not just my career journey, it is a mirror of what is possible for women in AI across Africa - if the right doors stay open.

I moved diagonally, gradually expanding my role and gaining experience in related higher-level skills. At the time, all I knew was that I was working, sometimes crawling, somtimes sprinting. I started as a Data Labeler, clicking through pixels and bounding boxes, delivering content moderation and other nlp use cases, teaching machines how to see and understand data. It did not feel revolutionary-it felt like survival. But with every annotation, I was learning the building blocks of AI–data.

When I became a Quality Assurance Lead, I realized how fragile the intelligence we were building was. One mislabel and the whole model could learn the wrong truth. As a Team Lead, I had honed my skills in precision and was entrusted to diligently manage multiple teams to project completions. Today I am a Product Associate, and I carry my formative experiences with me. I have lived the frontline, and now I translate that experience into how we build/train/finetune models, how we update our systems, and how we make sure our AI is not just functional–its fair.

Working in AI in Africa, sometimes feels like building a smart system on a potholed road–you are doing something cutting edge, but the basics aren’t smooth. We have a plethora of talent, we have data, and we have hunger. But we are constantly trying to catch up with infrastructure that was never built with us in mind. There are so many gaps. Gaps in monitoring, gaps in computing (grateful to Google StartUp Program), gaps in policy, gaps in pay. And the biggest one? Visibility. You can be training the model, leading the data team and still be invisible because your title does not sound technical enough for the global AI hype cycle. Everyone seems to be focused on finished projects, never the inner workings.

In every team I have worked on, it’s the women who keep the data flowing and the systems accountable. Yet we are rarely called engineers. I want to see more women break out of the annotation box and into decision making roles–not because they were lucky but because the blueprint was built for them to rise. And I want that future to include women who don’t code, who do not have PhDs, but who understand data intuitively, who see bias before the model does.

I have seen brilliant women from diverse works of life leave because no one told them their work was important. I have seen annotation guidelines written by people who have never labeled a dataset before. And I have seen how data that is not questioned becomes the bias that gets deployed.

Artificial Intelligence was never just technical–it is human in different ways. I share this not as a woman in AI, but as someone who believes that the future of machine learning isn’t just in code–it’s in the care, and women have always carried this important work.

---

Sharon Iphy is a Product Associate working in AI across West Africa. A Nigerian woman based in Ghana, she began her journey as a Data Labeler and grew into leadership, Quality Assurance, and product roles. She is passionate about building impartial AI systems, advocating for women in tech, and making the invisible work behind machine learning visible.