Technologies

​

React, Next.js, Python, LLMs, DynamoDB, Aurora, PostgreSQL

​

​

The Company

​

QuestionPro is a well-renowned cloud-based survey and research platform that empowers enterprises and individuals with essential tools for creating data-driven studies.

 

They boast over 5.3 million global users and are trusted by academic institutions, government agencies as well as tech giants such as Google, Microsoft and Amazon. 

Their platform additionally offers a comprehensive suite of tools for designing surveys, collecting responses through various channels such as email, social media, and websites, and analyzing data in real time.

​

​

What if the future can be predicted?

​

From product launches to discovering customer preferences, the QuestionPro platform was used to conduct frequent surveys and polls by the registered members.

 

This posed 3 main challenges: ​

​

• A large number of user responses had to be analyzed every time a new campaign took place. 

• It was time-consuming and caused the user subscription limits to be surpassed..

• Sometimes, the answers were too evident for previously asked, related-questions.

​

QuestionPro team required an AI-powered predictive solution to enhance their users' decision-making and reduce extra time spent on data analysis.

​

​

A product built to learn from the past

​

Confronting this challenge head-on, our team of engineers developed Synthetic - a scalable, web-based application. Synthetic was aimed at producing highly reliable responses based on user responses and patterns. 

​

Leveraging ChatGPT 3.5 and 4-Turbo, we implemented predictive capabilities into QuestionPro's survey tool. Based on the historical submissions of respondents to similar survey questions, users of QuestionPro's platform were able to get deeper insight into potential future responses to new survey questions. This allowed them to reduce the number of surveys conducted and depend on the predictions. 

​

During development, one of the principal concerns raised by QuestionPro’s team was the accuracy of the results. However, our team excelled at training the model to generate solid predictions. Compared to the actual responses received from the users, the AI-generated predictions were reliable. 

​

As a result, conducting high volumes of surveys was no longer necessary as synthetic takes care of it. Due to our value of clear communication with our partners, our team could deliver a state-of-the-art solution that was beyond exceptional. We still maintain a positive relationship and continue to provide ongoing development support. 

​

​

Client Testimonial

​

"Our collaboration with Paradigm was transformative, as their team demonstrated exceptional knowledge and expertise in AI applications. They played a key role in our strategic decisions, making them an invaluable partner in our journey towards digital innovation"

​

Vivek Bhaskaran
CEO/Founder at QuestionPro,
Texas, USA
 

​