While using a Business Intelligence tool, business users are often challenged with convoluted reporting use cases that require them to operate on complex data joins and aggregates. Let’s hypothetically say, an e-commerce giant’s Marketing team is interested in comparing the retention rate of users acquired by Facebook vs that acquired by Google. Getting this information would need the marketing team to model the data i.e. join, aggregate and summarize the data from multiple tables that store acquisition and retention data.
However, a surprisingly small number of BI tools allow building models on data without requiring to write complex SQL queries. This comes as a major drawback for business users who are not well equipped with the skills required to execute this. This puts the pressure back on the engineering teams to manually write code snippets that model the data and build scripts to move the result to the warehouse. Additionally, engineers end up spending considerable time & effort to monitor & maintain the entire system! Phew!
We, at Hevo, built ‘Models’ to absorb just this complexity.
In an attempt to fix the ever lingering problem of modeling data, Hevo Data brings a new feature that meets the unique needs of data folks, allowing them to create complex data models through a simple visual interface from within the Hevo Data Platform. This empowers the data teams to put forth a single source of truth to the whole organization thereby enabling business users to do more with their data.
With Hevo’s Modeling feature, any user can create complex joins and store the results as materialized views within data warehouses without having to worry about monitoring and maintenance. Business users can further build reports on these views, drill down, roll up and take data-driven decisions fast.
‘Models’ empower users to build complex data joins and aggregates on a beautiful visual interface. You can build your query, map the output to a table in your warehouse, and Voila! You are done. This makes data joins a breeze for the data heros. Some of the powerful features of ‘Models’ are:
Write, Preview and Commit: Easily build your query with the help of schema explorer. Avail a real-time preview of the join before committing the same onto the warehouse. Gives a major boost to the data team’s productivity, doesn’t it?
Build, Preview and Commit
Bird’s Eye View: Models are accompanied by a dashboard that provides an overview of all of the Models written on your Hevo Account. Detailed info about destination, status, frequency, etc. are all visible in one window for you to take action.
Details on all Models on your Account
Query History: Get a complete history of individual queries. Details on runtime, duration, status and output columns affected can be found. This is your go-to place to find the query’s chronicle of events.
Scheduling Control: Models and Workflows can be scheduled (Wow!), taking away the hassle of manually running them or writing additional scripts to automate schedule. Models can be scheduled to run at a specific time, a specific frequency or a combination of both. This enables the users to populate the model based on the team’s needs without having to break a sweat.
Setup Model Frequency
Workflows: Define dependencies between your models through Workflows. Workflows allow you to build a DAG (Directed Acyclic Graph) from within the tool, allowing you to optimize interdependencies between data models. Let’s go back to our initial example of extracting retention report by source. To achieve this, you will have to initially merge the data from the attribution table with the user table. You will then have to join this data with the user activity table, run aggregates on it to derive the final output. This 3 step process can be easily automated via drag and drop in Hevo’s Workflows. Cool, isn’t that?
Easy Workflow Setup
At Hevo, we are constantly striving to innovate and make a difference in your lives on a mega scale. ‘Models’ is a toddler’s step towards that grand vision.
What do you think of ‘Models and Workflows’? Let us know your thoughts and feedback in the comments section. We are all ears!
No-code Data Pipeline for your Data Warehouse