LiiRN: Machine Learning to Automate Change Management

Vivus created LiiRN's custom Machine Learning programming, to automate change management, using Python, Django, Celery, Javascript, Docker, and Tensorflow.

by Abby F., Aug. 10, 2020

Here at Vivus, our strength lies in our ability to perform all the legwork needed to make your digital product come to life. That’s why when LiiRN approached us with their idea to automate change management, we were excited to work with them to innovate the change management space. 

Vivus worked with LiiRN to automate change management which would enable enterprises to align everyone on a strategic initiative to accelerate transformation. This would build awareness and align decision-making with company-wide input to create a culture of change, from the bottom up. 

Machine Learning for LiiRN

Vivus built a custom web application that used machine learning to provide useful data to large companies. We focused on building a scalable Django application with APIs, task-management, and server-side caching. The result was an artificial intelligence driven platform that simplified implementing change throughout an entire enterprise with a blazing fast intelligent user interface and workflow to support interactions between varied clients and internal users.

Vivus built a robust machine learning production model that can be deployed anywhere in the world. Machine learning was incorporated using TensorFlow, an end-to-end open-source machine learning platform. Using TensorFlow gave LiiRN the ability to analyze text data such as questions, answers, and comments which then provided results of important keywords. This is important because LiiRN provides top companies such as Coca Cola important data on their annual change management assessment. Another benefit of going with TensorFlow is providing users sentiment analysis. Sentiment analysis is the interpretation and classification of emotions (positive, negative, and neutral) within text data using text analysis techniques. LiiRN analysts will only get better results with the more data the software receives. 

Repairing, scaling and new functionality

Vivus didn’t build the LiiRN application from the start. Our task was to repair any outstanding issues within the application, build new features including machine learning, and provide a custom scaling solution. Some of the tools we used were: 

  • Web Development using Python and JavaScript (Django/ReactJS). When we started the project, LiiRN was slow on both the backend and frontend. We provided support in both of these areas which lead to a massive decrease in page load time from 5.4 seconds to 2.1 seconds. We are still improving the load time as Vivus’ standard is 1.5 seconds page load time. 
  • API Development: Django REST Framework with full CRUD Functionality. We built a web API using Django REST framework. Our main focus was on processing time. LiiRN has clients with an employee count of 1,000+ meaning scalability and speed are a must. We used efficient queryset which allowed massive data retrieval and Redis for in-memory data storage with extremely fast response times that are less than a millisecond.


About LiiRN

LiiRN’s artificial intelligence-based software automates change management and measures behavior change to better align the entire enterprise towards achieving business transformation initiatives faster. Through Vivus’ custom machine learning web application, LiiRN has been able to accelerate brand transformation, change management, cultural transformation, digital transformation, diversity & inclusion, financial transformation, future of work, and market growth. 


About Vivus 

Vivus understands how to work with clients to turn leading technologies into high-quality finished products that solve real, ground-level business problems. 

Our dedicated experts here at Vivus would love to work with you to define, design, and build digital products. We offer custom software, mobile, and web application development and testing. Learn more about Vivus and what we can do for your organization at Vivus. 

Prev Project