AgriSmart DSS

AgriSmart DSS (Decision Support System) is a tech-enabled platform designed to help Nigerian farmers reduce post-harvest losses through smarter decision-making.

Comment

AgriSmart DSS — Reducing Post-Harvest Losses Through Data-Driven Decisions

Inspiration

Post-harvest losses (PHL) remain one of the most overlooked challenges in Nigeria’s agricultural sector. During a national hackathon focused on agricultural innovation, I was exposed to the shocking levels of food waste caused not by poor farming, but by lack of storage, inefficient logistics, and market disconnection. This inspired me to build AgriSmart DSS — a lightweight, data-driven solution that equips young farmers with the insights they need to reduce waste, cut losses, and increase profitability.

What it does

AgriSmart DSS is a decision support system that helps farmers:

  • Identify the nearest cold storage and market facilities.
  • Estimate transport and storage costs.
  • Understand optimal temperature ranges for different crops.
  • View regional spoilage patterns to aid in planning and investment.
  • Receive crop-specific guidance to reduce losses post-harvest.

The tool is tailored to smallholder farmers and agripreneurs using a mobile-first interface.

How we built it

  • Used Python and Streamlit to build an intuitive web interface.
  • Cleaned and structured a dataset of 1,000+ farm records including crops, locations, storage hubs, transport routes, and spoilage rates.
  • Designed a recommendation engine that filters and ranks facilities based on proximity, cost, and crop needs.
  • Implemented a lightweight knowledge base showing crop guidelines and investment priorities.
  • Created a fallback mechanism to suggest results even when inputs aren't a perfect match.

Challenges we ran into

  • Uneven team participation made coordination difficult; much of the technical work had to be carried out individually.
  • Initial MVP had poor documentation and analytical depth, resulting in a low score in a prior competition.
  • Data quality varied, requiring extensive preprocessing to standardize location names and crop inputs.
  • Feature creep — balancing between core functionality and bonus ideas like investment mapping or logistics integration.

Accomplishments that we're proud of

  • Delivered a functional MVP that addresses real pain points for Nigerian farmers.
  • Upgraded the original solution into a model-ready version with a more structured UI, logic, and output.
  • Developed a knowledge base that could be easily extended to support more crops, locations, or real-time APIs.
  • Designed for scale with future integration in mind (e.g., SMS alerts, AI-based spoilage prediction).

What we learned

  • The importance of grounding tech solutions in real-world user needs.
  • How to manage and visualize agricultural datasets for decision support.
  • How to build a user-facing MVP under time and resource constraints.
  • The need for early clarity in task delegation and project scope.

What's next for AgriSmart DSS

  • Tailoring the solution for submission to the Africa Deep Tech Challenge 2025, focusing on robust technical documentation and scalable architecture.
  • Exploring predictive modeling using spoilage, weather, and logistics data.
  • Integrating real-time price feeds and logistics providers.
  • Testing with real users in a pilot deployment and gathering feedback.
  • Building a companion SMS or USSD-based interface for farmers with limited internet access.

Built With

Share this project:

Updates