Inspiration
Post-harvest loss in Nigeria is not a statistic — it's a crisis with names, faces, and futures attached. Over $10 billion worth of food is lost annually due to systemic failures in logistics, storage, and market access. One of our co-founders witnessed a relative lose an entire season’s tomato harvest — not from poor yield, but from a buyer backing out and the produce rotting in the sun. Aevum Market was born from that pain: to solve a market failure that traps millions in poverty while food rots in abundance. We weren't inspired by AI. We were driven by injustice.
What it does
Aevum Market is an AI-powered digital platform that connects smallholder farmers directly to verified buyers before harvest, enabling secure, pre-arranged sales that reduce uncertainty, spoilage, and financial loss. Accessible via mobile app, USSD, and web, the platform ensures inclusion across literacy and connectivity levels. With escrow-backed payments, AI-driven demand and price forecasting, Llama-powered multilingual assistance, and optimized logistics planning, Aevum Market transforms fragmented agricultural value chains into transparent, efficient, and profitable systems for all actors involved.
Key Features:
- Pre-Harvest Linkage: Farmers get matched with verified buyers before harvest for planned sales
- Escrow Payments: Secures transactions and eliminates default risks
- Llama Voice Assistant: Supports users in Yoruba, Hausa, and Igbo via USSD and app
- Price & Demand Prediction: AI models guide planting and harvest timing
- Smart Logistics: Suggests optimal routes and cold storage points
- Traceability & Fraud Detection: Builds trust across the supply chai ## How we built it
How We Built It We began with ethnographic research and co-design sessions with 100+ farmers, aggregators, and food vendors across Oyo State. Our MVP combines: Frontend: React Native (App), USSD via Africa’s Talking Backend: Django + PostgreSQL ML Models: LSTM and XGBoost for price/demand forecasting; fine-tuned Meta LLaMA model for NLU across low-resource languages DevOps: Docker, GitHub Actions, Firebase Auth, integrated Twilio for SMS/voice We’ve deployed this stack to serve farmers with or without internet access, with a hybrid infrastructure that bridges the digital divide at scale.
Challenges we ran into
Connectivity gaps — solved with robust USSD and SMS-based support Digital exclusion — overcome by training 500+ farmers in basic digital literacy Trust deficit — tackled through escrow protection, verified buyers, and community ambassadors Language & literacy barriers — addressed with Llama-powered multilingual voice support Dynamic data environments — continuously feeding and retraining AI models with local agricultural datasets
Accomplishments that we're proud of
- 2nd place out of 466 teams across 54 African countries at the World Product Day Hackathon (Innovate Africa)
- $5,000 seed grant secured to develop MVP
- 500+ farmers onboarded in our pilot across 3 LGAs in Oyo State
- Partnered with 4 aggregation hubs and 2 logistics providers
- Reduced produce waste by 21% in pilot test cases
What we learned
Technology doesn’t scale on code alone, it scales on context. We learned to build for broken networks, to design for people who can’t read, and to deliver value on the first tap. We also learned that AI only matters when it’s human-centered — not just smart, but usable, equitable, and local.
What's next for Aevum Market
We’re ready to scale Aevum Market to 10,000 users in 12 months. We’ll deepen our Llama integration for dynamic decision support, extend to livestock aggregation, and expand across 4 Nigerian states. With support, we’ll convert a validated pilot into a national infrastructure for food resilience.
Log in or sign up for Devpost to join the conversation.