Segment AI - lead segmentation tool based on data science

Segment AI - Lead segmentation tool based on data science

About the product

Segment AI is a powerful marketing and sales tool. It uses data science algorithms to sort lead data into segments based on various criteria. As a result, it is possible to target each segment with special offers, increasing the efficiency of marketing and sales campaigns.

What the client came with

Our client had to deal with improper lead segmentation daily. He came with an idea of a product that could solve the problem for himself and thousands sales and marketing professionals worldwide. His decision was to use the power of data science to sort leads fast and exactly based on various available data.

Data science module

The product owner hired a team to build a Python-based module which was responsible for sorting lead data.

Web app design

Having certain expertise in web design, our client has created the user interface for the MVP version of his application

Clockwise Software’s contribution

The data science module was not enough to build a competitive product. It lacked an attractive frontend and a powerful and reliable backend to run properly. Our task was to build a strong system around the existing data science module. The process involved:

Frontend development

It was important to build a responsive application that would work on a variety of devices with different screen sizes. This was achieved with Bootstrap. We also had to consider how the sorting results would be displayed on the screen. React Bootstrap Datatables, Masonry and D3 were used to achieve different display modes.

Backend development

To build a reliable infrastructure around the core module, we used Laravel, a powerful PHP framework. Amazon Web Services, as one of the best cloud solutions currently available, was chosen to host the application. In addition, we implemented synchronization with Salesforce, and other major CRM systems follow.

Software testing

We extensively tested the whole system to ensure that the software had no major bugs that could affect user experience. In addition, automated testing was implemented to audit the functionality of the product.

Challenges of the development process

The process of creating a software product always meets challenges, especially when working with new technologies. Our development team willingly accepted these challenges and managed to release a competitive web application.

Limited time and resources

As the product development was the initiative of a sales and marketing enthusiast, the budget for the project was limited. It also wasn’t clear whether it was worth investing an impressive sum of money to build a full-feature application. So, the team agreed on MVP development, which allowed us to start testing the product with real users and release it in a short amount of time.

Complicated project setup

The project required complex initial settings, such as the installation of a specific version of Python, data science libraries, queues management tools Redis and Supervisor, backend framework Laravel/PHP and the database MySQL. Installing, setting up and testing in the new environment was a time-consuming process. There were also conflicting dependencies in the staging environment. To solve this problem, we used Docker containers.

Parallelisation and high consumption of computing resources

The data science module was performing complicated computations. As a result, the application had a slow response time. Our solution was to implement queues which allowed to separate computations into parallel processes. It significantly increased segmentation speed and as a result, offered better performance of the app.

Technologies and integrations

Backend:

Frontend:

Team and development time

Backend developer

Frontend developer

QA engineer

Project manager

3

Months

We will be pleased to hear from you, or receive a proposal for joint cooperation

Contact us