Cedar Valley Farms
Cedar Valley Farms is a project that helps cannabis operators increase their efficiency with the help of data-based insights.
About Project
1
Industry:
Cannabis cultivation
Type:
Data engineering
Time:
5 months
Idea
Our client is located in Colorado, US. The cannabis operators’ points of sale there usually provide their data to various services for storage and analysis. However, the different APIs that these services use make communication between them complicated.
The company wanted to facilitate the process of gathering and analyzing sales data from various locations, so they partnered with us to create a single database with analytical tools.
Business goals
Create a service that will collect and structure sales data into one database.
Facilitate data analysis by building visualization tools.
We were responsible for
Data pipeline
Authentication
Team
Project manager
Backend engineer / Data engineer
Industry:
Cannabis cultivation
Type:
Data engineering
Time:
5 months
How we built it
2
Discovery phase
We started the project with the initial research, business understanding, framing the problem, and defining requirements.
Data understanding
Here, we analyzed what data was already available for our client, determined the data properties, and chose the tech stack we were going to use to collect and process data.
Development
We have developed secure authentication, a data collection feature, and a three-step data pipeline that will allow dashboards and charts to use processed data.
Testing
We tested the solution and fixed some minor bugs.
Release
The model is released and the first users registered on the service.
Post-release support
We continue to improve and update the algorithms to ensure the most accurate data aggregation.
Key features
3
Data collection
The Cedar Valley Farms service collects data from analytical platforms once an hour and stores it as JSON files to prepare it for the data pipeline.
Data pipeline
A three-step pipeline includes:
Converting data into the canonical format to facilitate its analysis.
Data aggregation by various categories (location, brand, employees).
The set of REST API’s endpoints that allow using data in dashboards and charts (data frames).
When the pipeline is finished, users can access the data frames and take the necessary data.
JWT-based authentication
When users log in to the system, they receive a token with a certain level of access that allows them to see the corresponding data. It enables additional data protection.
Technology Stack
4
Challenge and Solution
5
Flowhub integration
Problem:
Flowhub is an analytical service that helps with managing data about cannabis sales. We needed to integrate it with our service, but Flowhub doesn’t have webhooks to set this up correctly.
Solution:
We used long polling to successfully integrate Flowhub into our platform.
What we have now
6
The project is released.
Business goals are achieved.
The service receives positive feedback from users.
