AI-powered questionnaire app  

About the project

To enhance engagement at a conference, we developed an interactive AI-powered questionnaire designed to showcase expertise in digital services, collect high-value contacts, and provide personalized offers. The goal was to create an engaging tool that not only educated users about the company’s capabilities but also seamlessly guided them toward services that matched their needs. 

Challenge

The questionnaire was built on a foundation of AI-driven personalization, using company-generated content such as website pages, blog articles, and case studies. Rather than presenting simple right-or-wrong answers, the system was meant to ensure that every response leads to a meaningful outcome, generating a customized service recommendation tailored to each participant’s business interests. At the end of the interaction, users were supposed to receive a personalized email summarizing their expressed needs and providing a targeted offer. 

Development process 

The project began with comprehensive data collection, aggregating over 100 questions derived from company materials. These questions were carefully designed to be evergreen and adaptable for different audiences. Instead of presenting a static quiz, the system dynamically selected a randomized set of 3–4 questions per user, ensuring variation and a more engaging experience. 

To refine the questionnaire’s effectiveness, we conducted extensive testing, iterating on question phrasing and logic to prevent bias or domain-specific limitations that could confuse users. This step was crucial in ensuring that the questionnaire remained intuitive and accessible while providing valuable insights. 

The next phase involved front-end and back-end development, integrating structured ChatGPT-powered AI prompts to analyze user responses. The AI system was trained to recognize user pain points based on their answers and generate tailored service recommendations that aligned with their expressed interests. By leveraging prompt augmentation, we optimized the AI’s ability to understand context, ensuring that the generated summaries and offers were both precise and highly relevant. 

Further refinement was achieved by implementing a complex structured prompt that blended text-based logic and JSON formatting. This approach allowed us to control multiple response variables, including tone of voice, content length, domain relevance, and customization levels. Since the questionnaire was accessed on mobile devices, we also established fixed text-length parameters to maximize readability and user experience. 

We also introduced a dynamic balancing mechanism to ensure equitable distribution of service recommendations. This prevented overrepresentation of certain offerings and provided a fair chance for users to explore a variety of company services. 

Technologies

React

LangChain

LanGraph

TypeScript

Python

Flask

Azure AI Search

Azure OpenAI

Azure AI Foundry

Business value

Beyond its initial use at the conference, the AI-powered questionnaire has become a valuable tool for ongoing lead generation. It streamlines the process of qualifying potential clients, offering them meaningful insights and bonuses, while simultaneously capturing their business needs. 

Additionally, the modular nature of the AI-driven framework enables continuous expansion, with the potential to integrate new content, refine question logic, and enhance customization capabilities in future iterations. 

By automating personalized outreach and eliminating manual lead qualification, this solution has not only improved user engagement but also strengthened the company’s ability to convert prospects into clients.