Designing AI for Real-World Use: UX Lessons from Diba Kaya

Artificial intelligence is transforming industries, but great AI isn’t just about powerful algorithms—it’s about how people actually use it.

Building a product that solves real user problems requires more than just innovation. It demands deep user understanding, intuitive design, and transparency—elements that many AI startups overlook.

To explore this in more depth, we spoke with Diba Kaya, an expert in UX, psychology, and human-computer interaction. Diba has an extensive background spanning cognitive and behavioural psychology, neuroscience, and industrial engineering. She has worked in big tech (Google, Amazon), healthcare, academia, and government—all with a focus on how humans interact with technology. Today, she is a Senior Researcher on the Search and Discovery team at JSTOR, where she leads product & monetisation research in the application of AI on the platform, search, recommendation models and teaching tools.

Diba shared game-changing insights on how AI-powered products can be designed to be intuitive, trustworthy, and seamlessly integrated into real-world workflows.

Here’s what every startup needs to know.

1. Context of Use: Where & How Will People Actually Use Your Product?

One of the most overlooked aspects of product design is context of use—the real-world conditions where people engage with a product.

Diba highlighted Segway as a perfect example.

💡 It had traction. It was innovative. But… where could you actually use it?

Was it for commuters? Tourists? Cities? Shopping malls?

🚶‍♂️ Without a clear use case, even great products struggle to scale.

How This Applies to Startups

For any startup, understanding context of use means asking:

• Where will users engage with the product?

• Will they be using it on mobile, desktop, or in meetings?

• Are they integrating it into their existing workflows, or does it require a major behaviour change?

📌 Lesson for Founders:

Context matters just as much as the technology itself.

Before scaling, ensure there’s a clear real-world use case for your product.

2. User Research: Talk to Users Before Scaling

Diba reinforced a key lesson that aligns with our recent conversation with Glasp.io’s founder Kei Wantabe—you can’t build in a vacuum.

User research isn’t just about validation; it’s about discovering how people truly interact with your product.

🔍 One of the best ways to enhance user research? Get more users on the beta.

🚀 More beta testers = more diverse use cases = deeper understanding of real pain points.

Why This Matters for Startups

Without enough user feedback, products can become too niche, too complex, or miss real user pain points.

Instead of guessing how users will interact with your product, founders should actively:

• Find consistent pain points (not just one-off requests).

• Ensure features fit multiple use cases rather than a single user’s workflow.

📌 Lesson for Founders:

More data = better product-market fit. Invest in beta testing, user interviews, and feedback loops.

3. Mapping User Journeys: Walk in Their Shoes

Diba emphasized the importance of visualizing the user journey rather than just thinking in features.

Instead of asking "what does our product do?" ask "what is the full experience of using it?"

Why This Matters for Startups

If users don’t understand how to interact with a product, they’ll abandon it.

Startups should focus on:

• Understanding how different user personas interact with the product.

• Identifying points of confusion and frustration.

• Ensuring every feature adds value, rather than complexity.

📌 Lesson for Founders:

Your product isn’t just a collection of features—it’s an experience. Map it out like a story.

4. AI Needs to Give Users Control: Trust Comes from Oversight

Diba highlighted an essential UX principle for AI: Users need to feel in control of AI-driven decisions.

💡 Why?

• If AI makes mistakes, users should have the power to intervene.

• Over-dependence on automation without oversight creates distrust.

How Startups Can Apply This

Startups using AI should ensure:

• Users can refine AI-generated insights.

• Clear explanations are provided on how recommendations are made.

• There’s an option to override AI-driven decisions.

📌 Lesson for Founders:

AI should empower users, not replace them. Give them control over the system, not just outputs.

5. Transparency: Users Need to Understand How AI Works

One of the biggest barriers to AI adoption is lack of transparency.

Diba emphasized that if users don’t understand how AI makes decisions, they won’t trust it.

🔍 Example: Google’s "Explainable AI"

Google’s AI tools now provide clear reasoning behind recommendations, making users more comfortable trusting automation.

How Startups Can Apply This

Founders should ensure their AI-driven products are:

• Providing a simple breakdown of how insights are generated.

• Allowing users to audit AI decisions.

• Making sure users understand what data is being used.

📌 Lesson for Founders:

Users are more likely to trust AI when they understand its decision-making process.

Final Thoughts: AI Should Enhance Human Workflows, Not Replace Them

From context of use to user control and transparency, Diba’s UX insights offer a clear message:

Want an AI-powered Data Agent that fits into your workflow?

➡️ Discover how Petals makes data insights effortless.