Summary of "Data Scientist from Sudan | Reem Elmahdi | Africans in Data Science and AI"
Background & Entry into AI
Rim Elmahdi is from Sudan and is now based in the UAE. She shares her path into AI/data science, her research background, and her current work—while also highlighting African ML initiatives such as Zindi.
Early education and transition
- Elmahdi studied Computer Science at Sudan University, then worked in software for a few years.
- She moved to South Africa for a master’s at the African Institute for Mathematical Sciences (AIMS).
- She then completed a research master’s in Applied Mathematics at Stellenbosch University, focused on prediction models for water-quality variables.
How she moved into data science/ML
Elmahdi says her move into data science/ML was largely by coincidence, sparked by a data analysis / machine learning course at AIMS. The course introduced her to:
- statistics
- big data
- machine learning
Adapting to Heavy Math
While AIMS profiles can suggest a pure-math focus, she explains that the institution is broader—not exclusively math/physics/CS—because students can choose modules that fit them.
What was hardest
- The biggest challenge wasn’t only the difficulty of the material, but the intense pace: large topics compressed into a few weeks with regular assignments and tests.
Confidence through strong peers
She emphasizes that learning alongside very strong peers at AIMS helped build her confidence, particularly by making it easier to:
- ask for help
- work through difficult topics until she mastered them
AI Communities & Africa-Focused Platforms
Zindi involvement (Sudan)
- Elmahdi is an ambassador for Zindi in Sudan.
- She participates in panels and outreach, including introducing Zindi to final-year students and fresh graduates.
Why Zindi stands out
She highlights several differentiators:
- It’s not just a time-bound event.
- It’s an ongoing platform with:
- competitions
- hackathons
- jobs
- blogs
- structured learning that’s “put into action.”
African datasets and problem-solving locally
She stresses that Zindi uses African datasets, arguing that local communities are best positioned to solve local problems.
Connection beyond Zindi: Indaba
She also supports Indaba, describing it as a platform that connects ML minds across Africa.
Example Projects & Current Work
Elmahdi works in the UAE at a company described as the DA Tech Lab, focusing on intelligent marketing.
ML/NLP tasks she works on
Her work uses machine learning/NLP for tasks such as:
- Emotion detection from social media posts about brands/products.
- Named Entity Recognition (NER) tailored to Arabic (especially GCC dialects), such as identifying locations/cities/country names within tweets.
A key research gap in Arabic NLP
She notes that there are fewer strong NLP models for Arabic compared with English, and frames the problem around choices such as:
- transfer learning from English to Arabic, or
- building models with added layers while accounting for linguistic structure and context differences.
Advice on Learning AI
Elmahdi advises learners to:
- avoid copying someone else’s exact path, since opportunities and learning styles differ
- recognize that students with CS/software backgrounds already have a strong problem-solving foundation for ML/data science
She recommends choosing a learning strategy:
- Top-down: build working applications first, then deepen theory
- Bottom-up: build math/statistics foundations first, then move toward articles and implementation
And ultimately: start.
Presenters / Contributors
- Host / Interviewer: Tanda (mentioned as “hi tanda” and asking questions)
- Guest: Reem (Rim) Elmahdi
Category
News and Commentary
Share this summary
Is the summary off?
If you think the summary is inaccurate, you can reprocess it with the latest model.