weeks / seasons
Turn a shifting season into clearer decisions.
Explore lightweight ways to translate forecasts into timing, risk and a few practical options — without another dashboard.
Independent studio · Lilongwe / Berlin
We explore how human-centred design, data and digital tools can support practical climate-smart agriculture.
Built for collaboration with
Working themes · 01—10
A working map of the questions we find important — from the timing of rain to the quality of partnerships. These are directions for inquiry, not finished answers.
weeks / seasons
Explore lightweight ways to translate forecasts into timing, risk and a few practical options — without another dashboard.
soil / resilience
Connect soil condition to everyday crop choices and longer-term field care.
field / catchment
Shared maps can reveal erosion, water paths and patterns that are difficult to see from the ground.
lost / kept
Bring local readings of dambos, slopes and riverbanks into practical landscape decisions.
yield / risk / taste
Make trade-offs around risk, taste, storage and habit tangible while respecting local knowledge.
give / return
Shape feedback loops where communities can work with patterns from their own observations.
phone / field
Explore roles for young digital stewards without defaulting to yet another app.
test / learn / adapt
Look for tools, routines and ways of organising work that can be adapted beyond one short project.
clear / steady
Make the working relationship as considered as the tool, study or programme it supports.
notes / atlas
Imagine an evolving view of where climate-smart practices are being tried in Malawi and what people are learning.
A practical collaboration cycle
A simple structure for shaping research, programme and product collaborations around real decisions.
Start with the decision people are trying to make, the conditions around it and the knowledge already in the room.
Translate research, local knowledge and field observations into a form that can be discussed, challenged and used without specialist language.
Keep the first version small enough to learn from, then adapt the tool or process with the people who use it.
A useful starting point
The best place to begin is a decision that feels stuck: information that is not reaching people, a pilot that needs a practical next form, or a field insight that deserves closer attention.
Tell us about the questionJournal · Research & field notes
Inclusive agricultural AI · 2026
Agricultural AI can exclude women, older farmers and people without smartphones. Evidence from East Africa shows how advisory services can be designed more inclusively.
Read the article