Coffee beans banner


We deploy artificial intelligence, machine learning and geospatial data to make research more effective.

Why Laterite

Drive for innovation

We are a bridge between established research methodologies and the latest advances in data technology.


At the forefront

We are continuously investing in our modeling, machine learning and geospatial analysis capabilities.


Out of the box

We offer data-driven solutions to commonly-found problems in socio-economic development research.


Predictive analytics

Predictive analytics and modeling

We use predictive analytic and modeling techniques to, for example, define target populations, extract insights from unseen data, extrapolate findings or plan ahead.

Some of the techniques we have used include: machine vision techniques to estimate coffee tree yields, Markov chain models to predict the evolution of education systems, gravity models to identify infrastructure gaps in Rwanda’s economic geography and economic complexity models to study the latent trajectory of economies.

Geospatial image jul 2022

Geospatial data

Laterite deploys geospatial data, such as terrain, climate, population composition and density, nightlights, distances from roads and urban centers, the vegetation index, to augment insights from survey data.

Adding insights from geospatial data brings two advantages: it strengthens the insights from econometric analysis and allows for those insights to be geographically contextualized. This, in our experience, leads to high quality spatial predictions on development-related indicators.

Ml ai

AI tools to accelerate research

Designing and conducting survey research takes a lot of time. Some of that time is spent on technical and repetitive tasks, like coding a survey, cleaning data or coding qualitative data. Laterite has developed a set of tools to answer these challenges, using the latest AI technologies.

We are currently in alpha mode testing and you will soon be able to check out our first batch of tools. (Subscribe to our newsletter to stay tuned.)