Redesigning Indigo Ag’s Legacy System for Efficient Land Sampling

After successfully implementing advanced software integration to reduce the data entry time, Indigo planned to continue with automation, specifically to streamline their land sampling process. The management of their legacy system required lots of manual work to do: employees had to visit fields, collect soil samples, send them to a lab for analysis, and then have data scientists interpret the test results after the lab returned them.

Additionally, the workflow was complicated by the need to secure access permits to private land and coordinate fieldwork with third-party contractors. To improve this process, a comprehensive redesign was essential.

For this redesign to take place, Intelliarts conducted due diligence of Indigo’s existing system. We identified its key limitations and presented a set of recommendations on how to improve it, including those connected to code refactoring and architecture redesign. We also developed a PoC to provide technical guidance on how to bring Indigo’s business idea to life.

Through the redesign, we minimized human intervention. Employees now use tablets to take photos and collect samples. We also implemented surveys tailored to specific fields and campaigns, which field agents fill out alongside taking photos. This feature enables the collection of extended data about the field and its environment, further enhancing the system’s capability to support R&D and analytics efforts.

The collected samples are then labeled with QR codes and sent directly to laboratories. Using QR codes, the system then automatically extracts the test results for our data scientists to process the data. Later on, this workflow was also enhanced by the installation of sensors on the fields, further automating data collection.

To tackle inefficiencies in sensor management, we replaced Indigo’s spreadsheet-based tracking system with a custom sensor inventory management system. Previously, a significant percentage of sensors were misplaced or lost due to manual tracking. With the new system in place, losses have been drastically reduced to fewer than 4 sensors per 1,000.

Key technological milestones that shaped the success of this project include:

• Hexagonal architecture: We chose a hexagonal architecture approach, which gave us flexibility in data storage and helped structure our solutions more efficiently. This architecture separates the core business logic from external dependencies, allowing for the switching of components like databases or APIs without impacting the core functionality. This improved scalability and maintainability, making Indigo’s system adaptable to future changes.
• Event sourcing: We utilized an event sourcing model to maintain a detailed history of inventory changes and sample data. This method captures every change as a sequence of events, enabling easy reconstruction of past states or rollback to previous versions if needed.
• Monorepo: The structure of the repository was built based on the principles of Monorepo, allowing multiple projects to coexist within a single codebase, sharing dependencies, tools, and configurations for greater operational efficiency.
• Microservices: Our services were built as microservices, aligned with the principles of 12-factor applications, ensuring they are scalable, maintainable, and easy to deploy across different environments without complex configurations.

One of the core challenges we faced was the lack of clear requirements from the customer so we had to consult Indigo on how to technically implement their business idea. However, the primary difficulty was rather organizational than technical because we had to bring together frontend, backend, and mobile to ensure seamless integration.

Today, we continue to collaborate with Indigo Ag, maintaining and supporting their mobile application after the successful redesign of their legacy system.

Our partnership with Indigo started from a standard request for software integrations but quickly transformed into implementing a wider functionality and lots of improvements to their tech ecosystem. Intelliarts also successfully built a long-term, trusted relationship with Indigo, and we look forward to continuing our collaboration in the future.

Here are some of the key business outcomes to highlight in the project:

• The redesign of Indigo’s legacy system contributed to automation, minimizing the need for manual input and eliminating inefficiencies caused by inconsistent processes. Earlier, key tasks related to land sampling relied heavily on individual knowledge, making these tasks time-consuming and difficult to scale. With the new system in place, workflows became standardized, transparent, and easier to manage.
• With a centralized and digitized approach, Indigo transitioned from relying on multiple spreadsheets managed by different individuals to a streamlined and efficient process orchestration. Now the team members can easily access and understand real-time data, reducing delays and enhancing decision-making. Data scientists have immediate access to lab results, accelerating the analysis phase and increasing productivity.
• The implementation of a sensor inventory management system reduced sensor losses from significant numbers to fewer than 4 per 1,000, ensuring efficient use of resources and reducing operational costs.
• During the system redesign, we also refined Indigo’s data analysis capabilities by addressing issues in the raw data collected by the legacy system. This improvement led to more accurate and actionable insights.
• Our successful partnership led to an ongoing relationship where Intelliarts now supports six interconnected services for Indigo. At one moment, we had five teams simultaneously working on Indigo’s project.
• Intelliarts engineers independently handled communication with third-party vendors, resolved data clarification issues, and prepared solutions and documentation. By taking over significant coordination and goal-setting responsibilities, the team reduced Indigo’s operational overhead, enabling them to focus on core business objectives.