Where intelligent systems are built
not just deployed.CARIS
The Center for Advanced Research in Intelligent Systems is WIATech's research arm — a working laboratory for AI, machine learning and applied systems, built to put Sierra Leone on the map of places that produce knowledge, not just consume it.
A research arm,
by founding intent.
CARIS is not an afterthought. It was written into the institute's founding charter alongside the academic programmes — because an institute without a working laboratory is not, in our view, an institute.
An institute that teaches intelligent systems must also build them. The classroom and the lab share a wall — and, often, a whiteboard.
CARIS — the Center for Advanced Research in Intelligent Systems — is the research arm of the Waterloo Institute of Advanced Technology. Its mandate, set out in the WIATech Founding Charter, is to pursue open-ended, long-horizon research on intelligent systems alongside applied projects in partnership with health, finance, agriculture and public infrastructure.
The lab is being stood up alongside the institute itself. It is in formation— recruiting faculty, finalising its first research lines, and preparing to receive its first student researchers from the AI & Machine Learning Track's capstone. This page describes CARIS as it is intended to be at full operation. The work begins as Cohort 01 begins.
From the WIATech Founding Charter· 2026 · § 04
Four lines of active work.
The first four research lines were chosen by the founding board for one reason: they are the places where serious AI work, done from Sierra Leone, would matter most. Each is described as drafted; refinements will follow as faculty are appointed.
- 01Research line
Foundation models for African languages.
Modern language AI is dominated by a handful of languages — and most of them are not ours. CARIS's first line of work is on foundation models for Krio, Mende, Temne, and other low-resource languages of West Africa — training, evaluation, and the data infrastructure that makes both tractable.
We expect to publish, to release datasets and benchmarks, and to ship usable models — not just papers — for the people who speak these languages.
- 02Research line
Applied Machine Learning for health & agriculture.
Decision-support systems for clinics and smallholder farms, built for the realities of West African infrastructure — intermittent connectivity, edge deployment, scarce labelled data, real-world stakes.
CARIS will partner with hospitals, ministries, NGOs and agritech operators to scope and ship work that is actually used — not pilot projects that die after a paper is published.
- 03Research line
Resource-efficient AI systems.
Model compression, quantisation, on-device inference, and energy-aware training. The engineering of AI that runs where the bandwidth, the compute, and the power do not.
Sierra Leone is a place where this matters every day. We believe the constraints of building here will produce work that the well-resourced world will eventually want too.
- 04Research line
AI safety, policy & society.
How intelligent systems should be governed in emerging economies — in dialogue with regulators, civil society, and the people they affect. CARIS's first contribution here is normative as much as technical: showing that AI policy questions deserve serious, locally-grounded research.
We expect to write public position papers, to advise on legislative drafts, and to host convenings that bring policymakers and practitioners into the same room.
When the work does begin —
this is what it looks like.
Photographs of CARIS at full operation, taken in the institute's lab. The empty room above is the place; this is the work. Both are needed for the same reason: an institute is the people who turn up to it.
Four principles
set in the charter.
Researchers have lots of ways to operate. The charter committed us to four. They constrain CARIS more than they enable it — which is, we think, the point.
Publish openly.
CARIS will publish to open venues and release datasets, models, and tools under permissive licences. Research that cannot be checked is, in our view, not research.
Ship things, not just papers.
Where the research line allows it, we will ship working systems — models, datasets, tooling, services — alongside the writing. The deliverable is more than the manuscript.
Small teams, real questions.
CARIS will operate with small, focused teamson questions chosen for their substance, not their fundability. We're trying to build a culture of doing the work, not justifying it.
Train the next layer.
Every research line will host student researchersfrom the AI & Machine Learning Track's capstone. Research is also a teaching responsibility — and a recruiting one.
The classroom
shares a wall with the lab.
CARIS isn't a separate institution. It is the same institute, doing different work in adjacent rooms. The lab feeds the classroom, and the classroom feeds the lab.
The people teaching you will also be the people doing the work.
CARIS faculty are the same people teaching the AI & Machine Learning Track. Office hours often double as working sessions.
The twelve-week capstone of the AI & Machine Learning Track happens inside a CARIS research line, supervised by a faculty researcher.
Where a capstone produces real work, it is offered the chance to continue as a CARIS collaboration after graduation.
Where we want
to work together.
CARIS is small and intentionally so. Partnerships are how the work multiplies — with funders, with industry, with peer institutions, with public-sector commissioners.
Commission a project.
A health ministry with a forecasting problem. An agritech operator with a dataset. A public-sector body with an AI policy question. CARIS scopes and ships.
caris@wiatech.edu.sl— 02Fund a research line.
CARIS welcomes foundations, academic funders, and corporate research sponsors who want to back open, applied AI work in West Africa. Multi-year support enables real depth.
Talk to us— 03Partner as a peer institution.
Universities, research labs, NGOs and think tanks who want a credible Sierra Leone-based partner on shared research, exchanges, or co-authored work. We're interested.
Introduce yourselfWhen the lab
publishes.
CARIS has not started publishing yet. When it does, the work — papers, datasets, models, write-ups — will go to one place, and to one email list. Be on it.
The CARIS
research dispatch.
A short, irregular email when the lab publishes something — a paper, a dataset, a research write-up, a position note. No marketing, no boilerplate. Unsubscribe at any time.
The first dispatch goes out when the lab has something real to share. We are aiming for late 2027— but we'd rather be quiet than premature.
Two ways
to step inside.
CARIS is recruiting. If you're a researcher who wants to work on something that matters — from somewhere that matters — read on.
Apply to lead
a research line.
CARIS is recruiting founding faculty across the four research lines. Looking for researchers with strong publication records, applied instincts, and a real interest in doing the work from Sierra Leone. Roles will be posted in Q3 2026; preliminary enquiries are warmly welcome now.
Write to caris@wiatech.edu.slApply to the
AI & Machine Learning Track.
The most direct path into CARIS as a student is the AI & Machine Learning Track — a three-year, 144-week flagship, with a twelve-week capstone inside one of the active research lines. The strongest capstones become continuing CARIS collaborations after graduation.
See the AI & Machine Learning Track