AI research: deep learning and new models
Understand, reproduce and advance AI models.
Your running project
Reproduce a paper on African languages
- The need
- Redo a published experiment, compare results, explain the differences.
- The data
- The open datasets of the chosen paper.
- External tools
- PyTorchHugging FaceGoogle Colab
Workshop
Establish the needs
Who uses it, what problem it solves, how to tell it worked.
What you will learn
- Linear algebra, probability and optimisation
- Neural networks with PyTorch
- Reading and reproducing scientific papers
- Running an experiment and writing up results
Certifications to aim for at the end
No certification opens the door to research: what counts is publications and a master's or PhD. These only attest to practice.
- Entry levelNVIDIA-Certified Associate: Generative AI LLM (NCA-GENL)NVIDIA
- IntermediateNVIDIA-Certified Professional: Generative AI LLMs (NCP-GENL)NVIDIAEnglish; two to three years' experience expected.
These certifications are issued by their own bodies, not by Nangmo, and their exams are paid for with them. Checked on 3 November 2026: confirm the language, price and exam location on their site before registering. These exams are generally taken without AI: you will need to read and reason on your own.
How to start a career in this field
- 1Strengthen your maths and build a small model with the guide, able to justify every line.
- 2Reproduce a recent paper and publish your code and your differences.
- 3Join an open research community, such as Masakhane on African languages.
- 4Prepare a master's or PhD; our Study Abroad service can help you build the application.
First jobs to aim for
- Research assistant
- AI PhD student
- Junior research engineer
The workshop is open, but this path's lessons are not online yet. Tell us you are interested: it helps us decide which to write first.