Real-Time Insight Augmentation

Boosting Qualitative Data Depth in Healthcare Market Research

Introduction

Our Client

Interrogare is a leading German market research institute founded in 1999, providing specialized insights for a range of industries, with a strong focus on the complex demands of pharmaceutical and healthcare clients.

Their Case

In high-stakes medical studies, free-text answers from physicians often lacked the necessary depth for robust analysis. Interrogare needed a way to elicit more detailed responses in real-time without disrupting the survey experience for these valuable experts.

Facts

Customer:

Interrogare

Services Provided:

ELICIT (by Ohja.ai)

Tech & Tools:

Google VertexAI, Python

Duration:

3 Months

Challenges

Insufficient Data Depth

A significant portion of open-ended text responses from medical experts were too brief or vague, limiting their analytical value.

High-Stakes Decision Risk

Shallow data from small, expert sample sizes posed a risk to the methodological rigor required for critical pharmaceutical development and market access strategies.

Seamless Respondent Experience

The solution had to integrate invisibly into existing questionnaires and generate medically accurate, context-aware questions to maintain credibility with physicians.

Approach

Ohja.ai developed a fully automated AI-powered service, accessible via Application Programming Interface (API), that integrates seamlessly into Interrogare’s digital questionnaires through simple HTTP calls.
The solution works autonomously, ensuring that every free-text response is instantly analyzed within the context of the research field. Whenever additional detail is needed, the system generates targeted follow-up questions designed to elicit richer, more actionable answers—perfectly aligned with Interrogare’s commitment to methodological rigor and innovative research methods.

Key Features:

Real-Time Analysis: Each submitted free-text response is immediately evaluated. Using a combination of advanced rule-based criteria (such as minimum text length or presence of keywords) together with state-of-the-art Large Language Models (LLMs), the service classifies physician responses as either sufficient or requiring follow-up.

Dynamic Follow-Up Generation: When a response is deemed to require follow-up, the API generates context-specific additional questions. These follow-ups are crafted to meet Interrogare’s stringent quality standards, ensuring that the collected data is both meaningful and usable for their healthcare clients.

Follow-Up on Equal Footing: A key requirement for physician acceptance was that follow-up questions be formulated with medical accuracy and professional relevance. The system therefore accounts for both the specific disease and medication being discussed, as well as the respondent’s own wording. Handling domain-specific terminology was essential, as general-purpose LLMs are not trained on specialized medical language.

Seamless Integration: The custom-built API was designed for effortless integration into Interrogare’s IT infrastructure. By leveraging a REST interface, the AI-based enhancement could be embedded directly into digital questionnaires, operating autonomously and delivering results within seconds—without manual intervention and fully aligned with Interrogare’s full-service capabilities.

Data Privacy & Compliance: All data processing and cloud hosting take place within the EU and strictly comply with European data protection regulations. Communications and data storage adhere to the highest security standards, particularly in cases where personally identifiable information (PII) may appear in free-text responses—an essential consideration for Interrogare as a healthcare market research institute.

Technical Implementation

The cloud service was designed to function independently and at scale:

REST API Integration: Interrogare’s questionnaires send real-time HTTP requests to the service. Each call processes a free-text response, returns a classification, and—if necessary—a tailored follow-up question.

Hybrid Analysis and Generative AI in Real Time: The system combines heuristic methods with LLM-based classification to assess response quality. Additional contextual data (such as domain-specific terminology) are incorporated to refine the analysis and guide the generation of follow-up questions. The generative component includes safeguards to prevent inappropriate output. To ensure a smooth survey experience, model selection and invocation were optimized so that even with follow-up generation, average response time remains below 1.5 seconds.

Cloud Deployment with Cost-Efficient Scalability: The service is hosted in “serverless” container deployments within the EU, ensuring GDPR compliance. This setup enables automatic scaling based on demand, with consumption-based pricing: additional service instances spin up as demand rises, and scale back during low-usage periods.

Outcome

Enhanced Data Quality: By automating both evaluation and enrichment of free-text responses, the service significantly increased the depth and relevance of collected data. According to Interrogare’s own analysis, in 90% of cases classified as requiring follow-up, the generated questions led to more meaningful physician responses.

Higher Yield of Valuable Responses: Improved response quality directly translates into stronger, more reliable research results. Interrogare’s pharmaceutical clients benefit from deeper insights across product development, communication strategies, and critical market access decisions.

Effortless Scalability: The fully automated, API-driven integration allows Interrogare to extend the solution across additional online studies without added complexity or intensive manual oversight—strengthening their position as an innovative research partner delivering modern, flexible, and tailored solutions.

Conclusion

Our collaboration with Interrogare demonstrates how a fully automated AI service can revolutionize data collection in healthcare market research. By embedding seamlessly into existing IT systems and operating autonomously, our solution not only improves the quality of free-text responses but also elevates overall data quality—delivering clear, actionable insights without compromising the user experience.
This case study highlights how AI can enhance methodological excellence and support customer-oriented research solutions.

Testimonial

The entire team at Ohja.ai not only supported us with AI from the very beginning, but truly inspired us. They don't just provide solutions; they dive deep into topics, understand, think along with you, and customize at its best. A very special project for us was the automated follow-up questions in online studies with physicians – this required not only understanding technical vocabulary and medical abbreviations, but also evaluating the substance of a statement in real-time and, if necessary, formulating a meaningful follow-up question. This feature is absolutely unique in the industry and impresses not only us but also our clients in the pharmaceutical sector—and delivers essential added value. Thank you Ohja.ai, we look forward to this great collaboration and to all the projects to come!

Stefanie Sonnenschein
Director Marketing & Communication
Interrogare GmbH

Bassam is our Visionary for AI-Powered Knowledge Scaling

With the precision of a scientist and the vision of a founder, Bassam pushes the boundaries of what's possible. He translates complex AI technology into tangible competitive advantages, especially in video automation. Talk to Bassam if you don't just want to optimize processes, but revolutionize entire business models.
Dr. Bassam Mokbel

Expert in AI and Machine Learning

Latest Work Highlights

Every challenge is unique, but the goal is always the same: to create transformative results. Here’s a glimpse into the types of solutions we've built.

Explore the Full Story

AI Content Co-Pilot

Scaling Thought Leadership for a Fast-Growing PR Agency

For Berlin-based PR agency Ampyre.co, scaling thought leadership content was a major bottleneck. We developed a custom AI-powered co-pilot that automates 80% of the manual work, enabling them to produce high-quality articles and posts at scale while maintaining full editorial control.

Explore the Full Story

AI-Powered Talent Sourcing

Unlocking the Passive Candidate Pool for a Manufacturing Leader

For manufacturing leader Jacob Rohrsysteme, high costs and slow hiring cycles were a major challenge. We built a proactive AI sourcing system that identifies and engages high-value "passive" candidates, resulting in an 87% response rate and a dramatic reduction in time-to-hire.

Explore the Full Story

AI-Powered Content Repurposing

Maximizing Reach for a Regional Economic Development Agency

For the regional development agency WEGE, manually repurposing content for different channels was a major drag on productivity. We built a custom AI content generator that automatically transforms a single article into tailored posts for social media and newsletters, quadrupling their output while retaining full editorial control.

Contact Us

Ready to explore how intelligent automation can transform your business?

Tell us about your biggest challenge or your boldest idea. We'll get back to you within one business day to schedule a conversation.

Topic

Message

Thank you for reaching out!
Speak to you soon :)
Oh, there was an error.