Past Projects

Our team boasts a strong track record of developing robust software solutions.

SurveyStack: HOME PAGE, APP, SOURCE

An open-source survey platform launched in 2020 and designed to empower shared community knowledge. The platform enables the creation of complex, custom surveys with special support for collection of geospatial and agricultural data. Users can send out surveys and collect submissions while controlling who can access the data across organizations.

SoilStack: HOME PAGE, APP, SOURCE, API

A progressive web app designed to support smart agricultural and environmental sampling. SoilStack has a robust API for sharing field boundaries and sampling locations, enabling seamless integration with external portals and MRV platforms. Soil samplers can use in-field navigation and data collection tools to guide their work. Since its launch in 2021, design-based sampling plans have been created for over 10,000 fields.

SoilStack Web Service: SOURCE

We developed a complex geospatial processing which ingests an area boundary, automatically fetches and clips publicly available data related to that boundary, re-projects and normalizes data from multiple data layers and runs sophisticated k-means and cLHS sampling algorithms to output a set of specific sampling locations.

Midwest Cover Crop Council: HOME PAGE

We were tasked with modernizing and enabling the long term maintainability of their cover crop tool which was originally developed and coded by students and scientists. We assessed the state of the tool and designed a solution to convert their PHP application into a Django application written in Python. We migrated their application utilizing security best practices, allowing for streamlined future development by student and scientist contributors, and providing a convenient admin interface for managing and publishing data.

Soil Carbon Prediction

Using reflection spectra and metadata we have developed classification and soil organic carbon prediction models for 4 countries. Models have ranged in complexity from simple classification models deployed as scripts within the SurveyStack platform that output soil classifications (degraded, average, above average) in real-time (offline compatible) to complex web services deployed in the cloud that take localized data (survey results), fetch geospatial data to match the sampling location and then output soil carbon percent values.

Data Visualizations and Dashboards

We developed an interactive visualization and reporting interface for producers to receive and explore results from the Bionutrient Institute. We have also built a prototype peer-to-peer benchmarking dashboard and custom reporting dashboards and data pipelines.