Using Azure Databricks to improve user engagement and retention

An educational publishing company approached Glorious Insights to help them with user-engagement and user-retention related challenges. Upon further discussion with their core team, our team of experts found out that it was a result of feeble data analytics and slow responsiveness. Here’s the whole case study for you where Glorious Insights helped them overcome all their business challenges and improve user engagement using Azure Databricks.

 

Customer background

One of the top three educational publishers provides services in learning math and science. It provides customized content in the form of text, software, and services. The company views education with a broader perspective. It takes education beyond classes in different formats and styles.

 

The company chose Glorious Insights to optimize the analysis of massive volumes of data related to students' and educators' engagement. Using a comprehensive analysis of this data the company wishes to develop products that can capture the learning process more accurately. It wants to deliver intuitive products that can be customized for individuals.

 

Challenges

The educational publishing giant relies on studying the data to understand the behavior of its users and improve engagement and retention. It uses massive datasets to learn about how students and educators are leveraging their products.

 

Its company continuously evolves to provide better learning management products for an enhanced learning experience. The company possesses enormous datasets with millions of anonymous records of students. It faces many challenges to process these data and optimize its processes.

 

  • Feeble data analytics

The company gathers its data from isolated sources. Its present processes are inefficient in analytics that causes the product development workflow to suffer.

 

  • Slow response time

Slower data pipelines increase the products' time to market. The delayed launch of new products reduces its promise and adversely impacts the returns.

 

  • Engaging new learners

The face of education and learning processes were changing at a great pace. This caused higher abundance levels from the students. The company's adaptive learning was unable to provide an improved experience for the learners.

 

 

Solution

Glorious Insights formed a strong team in collaboration with the company’s analytics team members. The search for an ideal platform that can deliver the required functionality with ease of implementation ended at Databricks unified data analytics platform.

 

We developed a new system and used the Databricks platform as a foundation. The system helps in making data-driven fruitful decisions in adapting and innovating the products.

 

  • Functional aspect

The new Databricks based system provides an interactive interface to the data science team. This enables the team to work coordinately using interactive notebooks and develop more relevant models. The data coming from various sources can now easily clusterized.

 

Features like auto-scaling empower the team with well-organized, easy to analyze information for improved operational efficiency. The system automated the company's workflow. This helps in building more reliable and faster pipelines to develop improved models for learning.

 

  • Technical aspect

The new system developed by us could reduce the infrastructural complexity to a remarkable extent. The Databricks technology provides a fully managed service on web service platforms like AWS. It reduces the cost of ownership and significantly simplifies infrastructure management.

 

The new Databricks based system provides impeccable data ingestion capability to allow data collection from heterogeneous sources. The ETL (extract transform load) process is highly efficient, thanks to the machine learning algorithms.

 

Results

The solution developed by Glorious Insights using Databricks helped the company in many ways. Its multi-fold benefits can be summarized as:

 

  • Deeper insight into data

The system could capture more detailed information about the students. It has already covered over ten million student interactions and has increased the company’s retention by as much as 19 percent.

 

  • Individually adaptive learning products

The system could better capture the comprehension level of students interacting with the company in some way. It helps in delivering more personalized recommendations to provide improved and result-oriented learning.

 

  • Rapid time to value

Data after the deployment of the new system suggest that the company could address more than 5.5 million students with its innovative solutions. This all happened more efficiently and much faster, increasing the time to value significantly.

 

  • Lower cost of ownership

With the new solution, the entire operation of the company is now on the cloud. This has taken down the operational cost by nearly 30 percent.