Explore the case studies below for examples of real-life solutions developed through our #AfricaBuildsAI hackathon.
Project: The goal of the project was to gather as much data as possible about embodied emissions and emissions during the lifetime of IT hardware and devices in order to help companies better track and manage their carbon footprint.
Outcome: The team created a data-driven solution to estimate the environmental footprint of various laptop models based on specifications and usage patterns. They created a comprehensive dataset, trained machine learning models to predict Global Warming Potential(GWP), and selected the best-performing model. The solution was deployed as a cloud API for easy user access.
Project: The goal of the project was to develop an Innovative Internal Platform API that allows job platforms and employers to better access Na’amal data in order to increase access to job opportunities for their trainees.
Outcome: The Na’amal project encompassed a series of essential steps, including data cleaning, migration, and hosting on DigitalOcean; the implementation of a periodic data sync with a cronjob; the development of a FastAPI-based API; and the addition of authentication. The API was meticulously documented to facilitate easy access, and all “queryable” attributes were listed.
Project: The goal of the project was to analyse gender differences in financing within the cassava value chain in Côte d’Ivoire in order to better understand the needs of women cooperatives and enable LadyAgri to better tailor their support.
Outcome: The team managed to conduct both statistical and text-based analyses on hundreds of survey responses. As well as the analyses themselves, the team developed a series of scripts to clean and pre-process the data. The insights uncovered were pivotal in helping LadyAgri design appropriate financial inclusion interventions for women cooperatives in the cassava value chain.
Project: The goal of the project was to understand patterns in user behaviour and build examples to enable predictive maintenance of Dance’s fleet of e-bikes. By analysing vast amounts of data, they aimed to uncover valuable insights that could optimise the performance and reliability of electric bicycles.
Outcome: The team developed data-driven descriptive analytics that will enable Dance to have a better understanding of how their e-bikes and batteries are used and maintained. The team also invested a significant amount of effort in cleaning and correcting the noisy data that was provided, analysed the bike miles, battery health, and level of charge, and then created visualisation figures and thorough interpretations.