2018 - 2020
OsoVega was an online marketplace where growers of speciality crops could sell their harvests to consumer-packaged-goods (CPG) companies. The platform prioritized helping growers obtain the best prices while helping CPG's gain access to inventory depth. In addition to purchasing crops participants were able to speculate on the future price of crops.
For the buyer side of the marketplace, a key problem is the need for a consistent supply at a predictable price. This is critical to producing products profitably and ultimately satisfying contracts with big retailers like Whole Foods.
For the growers, who were the sellers in the marketplace, two considerations became apparent. First, growers are conditioned to lock in a price for their crop based on the movement of futures prices. When the price is on the rise, or when the price-per-lb near grower's profit goal, they want to lock-in that price. Second, when growing a new crop, growers benefit from support with farming practices, yield-boosting farm inputs, and harvesting.
Osovega — Wireframes of the grower's sell-side dashboard and pricing management.
We opted for a hybrid e-commerce style interface that took inspiration from Amazon's seller site. Growers could list their products and set a fixed price or price range that they were willing to sell at. We staffed the marketplace with agents to assist in facilitating the transactions end-to-end.
In low volume markets, the spot price of goods can vary wildly, making it difficult to ensure a consistent supply. To address this challenge, we developed a farm production tool that helped growers - and the marketplace as a whole - predict the volume of crops likely to be available in a given growing season.
Osovega — Wireframe illustrating farm production tool with harvest prediction.
The production app tracked inputs and estimated yields based on feedback from specialists who visited the farms. This information helped growers plan their harvest and production schedules, as well as provide buyers with a more reliable estimate of future supply.
We chose to use C# DotNet and Postgres for the backend of the trading marketplace. The DotNet ecosystem for integrations and the language level compiler guarantees made sense in the context of this product. A series of Restful APIs provided the frontend access to the service.
For the farm production app we decided to use Python. Python's flexibility, speed and science libraries allowed us to deliver this part of the application quickly. It was also a breeze to integrate with weather stations, drone sensor data, among other data sources.
We chose a monolithic frontend application all written in React. This approach provided a seamless user experience for growers and allowed us to easily integrate marketplace data with production data coming from the fields.
© Spence Wetjen. Made in Austin, Texas USA.
