Automating Test Data
In the highly regulated cannabis industry, testing and compliance workflows are a constant challenge—varying by state and often reliant on manual processes. We created a streamlined solution to help cultivators collect, organize, and act on lab result data more efficiently, reducing bottlenecks and bringing more structure to a mission-critical part of the operation.
Client
Kiefa
Services
Design, Research
Industries
Cultivation
Date
January - July 2023
Impact
Thousands of test results processed — bringing structure to a previously chaotic workflow
Significant time and cost savings — automated ingestion replaced hours of manual labor
Became a key differentiator — quickly emerged as a top selling point for the sales team, driving adoption and growth
Kiefa's data parsing - we walked users through a quick review to double check all values and made it easy to spot by highlighting everything in pink.
Problem
Before products can move forward, cultivators must upload lab test results—a compliance-critical step that’s often manual, inconsistent, and costly.
Pain Points:
Manual workflows: Most rely on paper or spreadsheets, increasing errors and slowing down operations.
Inconsistent lab formats: Varying data structures make standardization a challenge.
High cost of management: Testing is expensive—and so is the time spent organizing results.
Lack of insights: Unstructured data limits opportunities to track trends or improve quality.
Impact Areas:
Boost operational efficiency
Improve compliance accuracy and speed
Drive quality improvements through data
Enable scalability without added overhead
Product Opportunity
How might we automate lab data ingestion, standardize test formats, and turn compliance into a strategic advantage?
An example of some varying formats that operators need to parse through for the same information
Solution
We started with a manual data entry MVP to validate the need for structured test result management and gather real-world usage patterns.
Key Features & Outcomes
Visual Analytics: Surfaced trends and averages across harvests, giving operators better visibility into product performance.
Strain Comparisons: Introduced a dedicated view for cross-strain analysis—enabling marketing and sales teams to differentiate offerings and attracting new user segments.
Compliance Integration: Automated test result uploads via a third-party API, significantly reducing manual work.
v2: Automating Ingestion
To eliminate the biggest friction point—manual data entry—we built an intelligent parsing engine:
AWS Textract to extract tables from lab result PDFs
ChatGPT to interpret structure and isolate cannabinoid/terpene values
Levenshtein distance to match extracted compounds against our database, improving accuracy and reliability
This evolution moved the product from a reactive compliance tool to a proactive data engine—unlocking insights while reducing overhead.
This page houses all COAs that have ben uploaded as well as allows users to review the data we’ve extracted.
Under the "Test Results" section, users see the total averages based on all uploaded test results. They can click on a specific value to view changes over time.
This feature allows users to select specific strains to compare and contrast specific test result values.
A very valuable aspect of collecting all of this data was the ability to export historical reporting to send to sales, marketing and higher level decision makers that wouldn't regularly be on the tool.