From 0 to 1: Scaling Customer Research to Validate and Build Faster—My Journey with Generative AI in a scale-up??

As a recent Head of Product at Skydreams, a fast-scaling company operating in 20 markets, I’ve had to adapt to a fast-paced environment where speed and execution are critical. Of course, customer discovery and research are key—at Skydreams, we believe that a great product drives business. But in a rapidly scaling company, moving too slowly on research can mean losing momentum.

Traditional user research methods are valuable but can be time-intensive and resource-heavy. When I joined Skydreams, I needed to rethink how we extract insights, validate ideas, and execute quickly without compromising on customer understanding.

This is not a post about how to do customer research; rather, I want to share how I’ve leveraged Generative AI to accelerate insights, uncover pain points, and de-risk product decisions—especially useful for teams looking to optimize their research process, overcome budget constraints, or explore AI-driven approaches in product operations.

Here’s how I’ve leveraged Generative AI to turn constraints into opportunities:

?? 1. Research Planning: AI as Your Research Strategist

Research planning can be slow, often centralized in a small team or even one person, making it difficult to scale effectively. This bottleneck delays decision-making and impacts time-to-market, especially in fast-paced environments where speed is critical.

?? What Are Custom GPTs? OpenAI introduced Custom GPTs in late 2023, allowing users to configure AI models with domain-specific knowledge and tailored instructions. These models enhance productivity by automating tasks that traditionally require significant manual effort and expertise.

?? How I Configured My Custom GPT for Research Planning: To create a GPT that could help me craft structured research plans, I configured it with specific instructions to function as a UX research expert specializing in marketplace businesses like Getamover.co.uk. My goal was to make research planning more structured, scalable, and data-driven.

I structured the model with:

• A defined role: The GPT acts as a UX research strategist focused on marketplace dynamics and lead generation.
• Research methodology guidance: It determines the best research approach, whether exploratory (interviews, benchmarking), validation (surveys, A/B testing), or measurement (behavioral analytics).
• Business context and focus areas: The GPT understands Getamover’s B2B and B2C needs, helping prioritize research on lead quality, pricing models, and form optimization.
• An adaptive research framework: It prompts for critical elements such as research objectives, hypotheses, expected outcomes, and business impact before generating structured research plans.

?? Real impact: For a new feature targeting service professionals, I leveraged AI to:

• Crafted a thorough research plan covering both B2B and B2C perspectives.
• Structured the plan into Generative and Evaluative Research Phases, defining research methods, rationale, and expected outcomes.
• Reduced research planning time from 2 weeks to 1 day, followed by an additional day for iteration and stakeholder alignment.

AI doesn’t replace our product strategy—it strengthens it, giving us the ability to focus on critical analysis, deeper insights, and impactful decision-making rather than getting stuck in repetitive groundwork.

For other references, take a look at this NN article.

?? 2. Generative Research: AI-Powered Qualitative Research at Scale

Conducting qualitative research is time-intensive and resource-heavy. Typically, researchers are fortunate to interview around 20 participants over a span of 2-3 weeks due to the effort involved in scheduling, moderating, and synthesizing results. When you introduce variables like multiple languages, time zones, and diverse regional markets, the interviewing phase alone can drain significant resources. Scaling research efficiently without sacrificing quality remains a major challenge, making it difficult to move fast while ensuring depth and accuracy in insights.

?? What is Tellet.ai? Tellet.ai is an AI-powered research tool that conducts large-scale, adaptive qualitative interviews via chat, voice or video in multiple languages, reaching multiple participants simultaneously. The platform automates the interviewing process by dynamically adjusting follow-up questions based on user responses, ensuring richer insights while reducing the burden on research teams. While it doesn’t replace the depth of a traditional qualitative interview, it enables efficient large-scale user research.

?? Using AI to Scale Interviews I leveraged Tellet.ai to interview our service professionals efficiently. Like Custom GPTs, Tellet.ai requires structured configurations, including context, research goals, user segments, and a questionnaire guideline to function optimally. The experience is highly conversational, ensuring that interviews feel natural rather than robotic.

Once interviews were conducted, Tellet.ai synthesized large volumes of qualitative data by:

• Grouping responses by themes and urgency, enabling faster pattern recognition.
• Highlighting emotional sentiment trends, helping prioritize issues based on user frustration or enthusiasm.
• Answering business-specific questions: The AI can be prompted to analyze responses and provide answers to specific business inquiries based on participant feedback.

While AI-assisted analysis is powerful, I found that these features are still evolving. To ensure accuracy and relevance, I performed my own analysis to extract and prioritize insights effectively.

?? Real impact: For a recent initiative, I used Tellet.ai to:

• Conducted 80 customer interviews asynchronously across 5 countries in 2 days, inviting participants via email to enter a chat-based AI interview, allowing them to respond at their convenience.
• Captured 25 minutes of in-depth feedback per participant via AI-driven chat-like interviews, ensuring rich qualitative insights.
• Uncovered key insights that directly shaped our 2025 product strategy, driving data-informed decision-making for upcoming initiatives.

While Tellet.ai doesn’t fully replace the depth of traditional qualitative interviews, it dramatically improves efficiency, allowing for rapid user research at scale while maintaining high-quality insights.

?? 3. Evaluative Research: AI-Driven UX Testing & Prototyping

User testing and concept validation are critical but often slow and resource-intensive. Prototyping, designing wireframes, and gathering structured user feedback can take weeks, delaying decision-making and product iteration. Additionally, iterating on UX concepts before committing to full design and development requires significant effort from design teams.

?? Using AI for UX Ideation & Prototyping: I’ve been experimenting with tools like Galileo AI and UX Pilot AI, which generate UI/UX wireframes and design flows based on text prompts or sketches. While still exploratory, this approach enables:

• Faster concept visualization: AI-generated designs help quickly shape ideas without requiring immediate designer involvement, and by incorporating our design system and brand guidelines, AI-generated components become more aligned with our existing UX patterns.
• Iterative design exploration: Tools like UXPilot AI generate three different design variations per prompt, allowing you to choose the most suitable version and refine it further with additional prompts.
• Seamless integration: These designs can be exported directly to Figma, making collaboration with design teams more efficient.

?? Extending AI’s Capabilities for Future Testing: While I haven’t yet used AI-generated designs for user testing, it’s something we are actively exploring. Tools like Builder.io go even further by turning AI-generated designs into interactive coded prototypes, which could enable faster usability testing in the future.

?? Impact:

• Accelerated UX ideation: AI-generated design concepts help streamline discussions and early-stage prototyping.
• Reduced dependency on manual design iteration: Experimenting with AI for rapid mockups saves time before committing design resources.
• Future potential for automated usability testing: We are evaluating AI-generated prototypes as a way to collect user feedback faster.

Final Thoughts: The Fast Evolution of AI in UX Research

The rise of generative AI in UX research isn’t just about speed or efficiency—it’s about redefining what’s possible. As I reflect on my journey at Skydreams, I’m convinced that the future belongs to leaders who embrace AI as a collaborator, not just a tool. Here’s what I’ve learned:

1?? AI amplifies ambition, but humans own the vision. Tools like Custom GPTs and Tellet.ai let us scale research across 20 markets, but it’s our team’s curiosity that turns data into strategy. 2?? The “art of the prompt” is the new leadership skill. Precision in guiding AI—whether structuring research plans or refining prototypes—determines the quality of outcomes. 3?? Move fast, but stay humble. AI accelerates validation, but it’s our responsibility to question biases, fill gaps, and stay grounded in real human needs.

My biggest takeaway? AI doesn’t eliminate constraints—it reshapes them. We’re no longer limited by bandwidth or borders; we’re limited only by our willingness to experiment, adapt, and learn.

To fellow product leaders: The next frontier isn’t about replacing people with algorithms. It’s about building teams where AI handles the “what” and “how,” while we focus on the “why.” At Skydreams, this mindset has let us innovate faster, de-risk boldly, and stay ruthlessly customer-centric—even at scale.