Building open-source AI solutions for agriculture and sustainability.
In-person hackathon in Nairobi focused on climate and market resilience. Teams built solutions to help anticipate climate shocks and market volatility, with the goal of protecting smallholder farmers’ incomes and improving decision-making under uncertainty.
Dates: 2025-12-03 — 2025-12-05 | Location: Impact Hub Nairobi, Nairobi, Kenya
Participants: 126 | Teams: 0 | Models produced: 3
In-person hackathon in Nairobi focused on inclusive credit scoring for smallholder farmers. Teams developed data-driven approaches to assess creditworthiness for farmers with limited formal credit history, aiming to unlock access to agricultural finance and improve financial inclusion.
Dates: 2025-10-29 — 2025-10-31 | Location: The Piano, ALX Hub, Nairobi, Kenya
Participants: 91 | Teams: 0 | Models produced: 3
In-person hackathon in Accra bringing together participants to develop customer segmentation and retention intelligence for agri-ecommerce. Teams worked on applied data solutions to help improve user targeting, increase retention, and drive sustainable growth for agri-focused marketplaces.
Dates: 2025-10-24 — 2025-10-25 | Location: Academic City University College, Accra, Ghana
Participants: 50 | Teams: 0 | Models produced: 3
Online Zindi challenge with Rhea to estimate soil nutrient levels when laboratory testing isn’t available. Using geospatial and soil data (plus optional earth-observation signals), participants built regression models to predict multiple nutrients—supporting soil-health assessment and tailored recommendations for more sustainable, resilient farming.
Dates: 2026-02-04 — 2026-03-06 | Location: Online, Online, Kenya
Participants: 896 | Teams: 0 | Models produced: 3
Online Zindi challenge with DigiCow to predict whether farmers will adopt improved practices after training. Participants built classification models that estimate adoption probability using only information available at training time—enabling smarter follow-ups, tailored support, and stronger extension programme design.
Dates: 2026-01-23 — 2026-02-28 | Location: Online, Online, Kenya
Participants: 896 | Teams: 0 | Models produced: 3
Online Zindi challenge with Farm to Feed to build a recommender system for fresh produce e-commerce. Using anonymised transaction data, participants predicted purchase likelihood (7 and 14 days) and expected quantities per customer and SKU—helping move surplus produce, reduce food waste, and expand market access for smallholder farmers while improving operations and marketing.
Dates: 2025-11-28 — 2025-12-28 | Location: Online, Online, Kenya
Participants: 733 | Teams: 0 | Models produced: 3
Online Zindi challenge with agriBORA to forecast weekly maize prices across key Kenyan counties. Participants built machine-learning time-series models to predict average weekly prices, helping farmers and cooperatives time sales, reduce post-harvest losses, and strengthen agriBORA’s warehousing, credit, and market-intelligence services.
Dates: 2025-11-14 — 2025-12-28 | Location: Online, Online, Kenya
Participants: 1095 | Teams: 0 | Models produced: 3
Online Zindi challenge co-led with Tolbi to build a pixel-level crop classification model from open-source Sentinel-2 satellite imagery. Participants developed machine-learning approaches to distinguish cocoa, rubber, and oil palm plantations—enabling scalable crop mapping for smarter land-use decisions and precision agriculture in Côte d’Ivoire and beyond.
Dates: 2025-04-24 — 2025-06-16 | Location: Online, Online, Ivory Coast
Participants: 866 | Teams: 0 | Models produced: 3