guhuza (Kinyarwanda): to connect
SEESALT Terra · Kigali and Rwamagana, Rwanda
A card game that gets young people talking about what nobody says out loud.
Rwandan young people can recite the facts about HIV. Saying the word in a room of their peers is something else. Guhuza is a deck of 108 cards made for that distance — played in groups, on paper, with no electricity, no internet and no clinician in the room.
Knowing is not saying
The barrier is not what people know. It is what happens when they ask.
Health education in Rwanda has done its work. Young people can name the transmission routes and the testing intervals. What keeps them from a clinic is not missing knowledge but the social cost of being seen to need it — something felt in the body, in a room full of peers, before anyone puts it into words.
A game moves that weight. On the table, a difficult word belongs to the card that asked for it, not to the person who says it aloud. That small displacement gives someone a reason to speak in front of their peers, and a shared script for what comes next.
The timing matters. PEPFAR financed 37.7 per cent of Rwanda's HIV response and is withdrawing. What survives that has to run without a budget line: on paper, in an ordinary classroom, led by someone who already lives there.
July 2025
Forty decks, four institutions, and nobody flown in to run them.
Roughly 330 participants across Kigali and Rwamagana. Rwandan student leaders facilitated every session. When travel risk rose we moved delivery forward a month; when scale started to strain the sessions we wrote sign-in and role-assignment procedures mid-pilot rather than after it.
Partners kept their decks, and three people asked us back: ULove, the mental-health club at African Leadership University; an ALU researcher working on HIV who wants to join the project; and a teacher at Agahozo-Shalom. None of them asked for more cards. They asked for more sessions. The deck mattered less than the gathering it made possible, and that finding reorganised the project: what is scarce is people trained to hold the conversation, not cards.
Content was co-designed with Partners In Health staff, a Rwandan development consultant, and faculty at the Middlebury Institute of International Studies. Production was funded by a $10,000 Projects for Peace grant.
- African Leadership University15 decks. ~300 undergraduates. Validated the facilitation mechanic; now supplies trained student facilitators. Its ULove mental-health club has asked for a mental-health module.
- Agahozo-Shalom Youth Village15 decks. ~30 students at a residential secondary school enrolling 500+ orphaned and vulnerable young people from all 30 Rwandan districts — a concentrated national sample of exactly the young people formal services miss.
- Partners In Health Rwanda5 decks. Co-designed the clinical content of the first module.
- Africa Edu5 decks. Distribution partner for the pilot placement.
Forty of the fifty printed decks were placed; ten are held in reserve.
How it works
Co-design the content, train local facilitators, hand over the deck, leave.
The toolkit is designed to be abandoned by us. A deck that needs us present is a deck that stops working the moment funding does.
What is in the box
- 108 cards across four content domains
- Every card in English and Kinyarwanda
- A 20-page facilitator's guide
- Session procedures for sign-in and role assignment, written during the pilot rather than before it
One architecture, many topics
HIV is the first module. Mental health is next — symptom recognition, help-seeking, and the stigma attached to both. ULove, the mental-health club at African Leadership University, asked for it directly.
Mental health asks more of a card than HIV did: much of what it deals with is felt before it is named. We want to build it with people who already work in that territory, not for them.
The structure does not change when the subject does. That is the reason to build a deck rather than a campaign: the second topic costs a content cycle, not a new product.
When someone says more than the game asked
A game that makes it easier to speak will sometimes let someone say something the room is not equipped to hold. The mental-health module will not go into a room until it has a written protocol for that moment: what the facilitator says, what stays in the room and what cannot, and which named service nearby the person is referred to.
We will write that protocol with counsellors who already do this work in Rwanda, and publish it on this page.
Two languages
Every card exists in English and in Kinyarwanda, translated by Esther Kwizera Urindwa. The choice is not cosmetic. A difficult word can be easier to say in a second language, held at a little distance, or harder, because it sounds like a classroom rather than a home.
Which language a room reaches for, and when it switches, is something we want to learn from the facilitators rather than decide for them.
One session
Players sign in and are given roles before the first card is drawn — procedures written during the pilot, when growing numbers began to strain the sessions.
Who runs it
Coordinated from New Haven and New York. Everything that matters happens in Rwanda.
- Zhijun He — Co-DirectorMA candidate in African Studies at Yale. Partnerships and policy engagement.
- Xinyue Wen — Co-DirectorPhD student at the Columbia School of Social Work. Product, grounded in her research on spatial inequality and population health.
- Mengyao Zhao — Tsai CITY RepresentativeMPH candidate in Social and Behavioral Sciences at the Yale School of Public Health. Represents Guhuza in the Tsai CITY Launch Pad.
- Esther Kwizera UrindwaLocalisation. English–Kinyarwanda translation of every card.
- Lisa AkamanziOperations, and the Rwandan entity now being structured.
- Cynthia Umugwaneza, Vanessa Munyana and EricFacilitation and outreach.
The players are not the buyers. Decks are bought by schools, youth NGOs and donor-funded health programmes that already hold budgets for teaching materials.
Modelled landed cost of one deck in Rwanda at a 500-unit run, including EAC tariffs — around $4 ex-factory. One deck serves three to six players across dozens of sessions, which puts material cost per participant in cents.
Guhuza is pre-revenue. These are modelled unit economics at volume, not realised ones. Product development is already amortised: the Projects for Peace grant covered content adaptation, artwork, the facilitator's guide and the first 50 decks.
Revenue is planned in three tiers — a print-ready licence with facilitator training, finished decks, and co-branded editions for NGOs. Decks are a one-time purchase; training, new modules and content updates recur.
What the numbers do not tell us
The numbers count people and decks. They stop when the session ends.
Every figure on this page is a count: participants, decks, requests to return. None of them tells us what happened after a session ended — whether someone who said a word aloud went on to test, to ask for help, or to keep talking outside the game.
That is the gap that matters. Saying is not the same as doing, and a game that makes speech easier has done only half of what it is for if nothing follows. Finding out what follows, by staying with facilitators and players over time rather than counting them, is the next piece of work.
Our own position is part of the picture. The project is coordinated from the United States and the sessions happen without us. What we know about them comes largely through the people who ran them, which is the design working as intended, and also a limit on what we can claim.
What's next
A request for more sessions is a request for people, not things.
Fall 2026 · Launch Pad venture
Guhuza has been selected for the Fall 2026 Launch Pad at the Tsai Center for Innovative Thinking at Yale — a multi-week incubator that takes early-stage ventures through skill-building workshops and mentorship alongside a cohort of student founders. Mengyao Zhao represents Guhuza in the cohort.
The semester goes to the part a pilot cannot settle. We have demand we cannot yet transact against: no Rwandan entity that can invoice the institutions behind those requests, and no price we have actually tested. Registering locally or finding a fiscal host comes first, then paid pilots with the partners in hand.
Running alongside that is a second edition of the deck, with its own mechanics and artwork, and a scoring design adapted for questions that have no right answer.
If you work on adolescent health, stigma, mental health or youth programming in East Africa, as a practitioner or as a researcher, we would like to hear from you. Write to us through SEESALT.
播种光明，收获永恒
© 2026 SEESALT · 世光 · Terra · About Terra · SEESALT home