Cohort 9 - Flipped

Cohort 9 was the first to run in a fully flipped-classroom format. Participants watched curated pre-recorded lectures and worked through labs during the week, then met every Saturday for a 2-hour live community session via Zoom. Daily interaction, Q&A, and accountability happened on Discord.

What participants did

Each week, participants were:

• Assigned selected videos from a curated playlist of lectures and labs
• Given supporting materials - Jupyter notebooks, slides, assessments
• Joined a live Zoom session on Saturdays to engage with instructors and peers
• Interacted daily on Discord for Q&A, collaboration, and accountability

By the end of the cohort, every participant either completed a capstone project as part of a team and presented their solution, or earned a certificate of completion based on the cohort’s attendance and assessment requirements.

Curriculum at a glance

A 15-week programme: 10 weeks of structured learning followed by 5 weeks of capstone project work.

Week	Topic
1	Python & numerical computing
2	Data science foundations
3	Databases, SQL & exploratory data analysis
4	Math for ML
5	Text processing
6	Linear regression & classification models
7	Non-linear modeling & interpretable ML
8	Probabilistic models
9	Unsupervised learning & recommendation systems
10	Deep learning basics
11–14	Capstone project work, mock presentations, final demos

Lectures and labs are publicly available on the cohort’s YouTube playlist. All materials, assignments, and resources live in the cohort GitHub repository.

Capstone projects

Ten teams shipped capstone projects spanning prediction, classification, recommendation, and risk modelling.

Team	Project
BenOkri	Predicting customer churn in Nigeria’s telecom industry - an ML approach with MTN data
Armah	Research paper clustering for topic discovery and literature organisation
SeedGuard	SeedGuard AI - mapping the GMO narrative landscape for agricultural sovereignty
TrulyFit	Building a rich fitness dataset and recommendation system
Nwapa	Predicting solar energy efficiency of buildings in Lagos
SDML	Spam detection in emails using NLP
Adichie	Flood area extent prediction in Ibadan metropolis using GIS and ML
TwinPillars	Predicting cholera risk in Kenya and Nigeria - a data-driven approach to preventive public health
Gordimer	Nutritional value estimator for Nigerian foods
Solo-Alie	Customer segmentation with RFM and unsupervised learning

Certification

To receive a Certificate of Completion, participants needed:

• 60% minimum attendance at community calls (tracked via Google Forms)
• 40% average assessment score
• 100% participation in the final project (submission required)

Acknowledgements

Cohort 9 builds on the foundation of Cohort 8 - its lectures, labs, and community contributions remain the structural spine of every cohort that follows. Deep gratitude to the volunteer instructors, lab facilitators, mentors, and organisers who continue to make this work possible.

Credits

Instructors

Kenechi Dukor · Oluwafemi Azeez · Tejumade Afonja

Organising team

Jesuyanmife Egbewale (lead) · Tejumade Afonja (co-lead) · Adetola Adetunji · Ibrahim Gana · Sharon Alawode · Simon Ubi