Product Designer, UX Researcher
2 weeks
Research, IXD, Visual Design
1 PM, 1 Eng, 1 PD
The idea was simple: an AI agent to automate routine tasks and enhance efficiency to be hosted on our partners' sites.
The reality? Skepticism, confusion, and lack of trust from users. While AI was a hot topic, partners weren’t convinced it could actually improve their workflows.
They feared losing control, questioned accuracy, and saw it as “just another tool” rather than a solution to their needs.
The challenge wasn’t just launching an AI feature—it was proving its value in a way that made users actually want to adopt it. Without high adoption, the AI agent would become yet another underutilized tech initiative.
65%
70% QoQ
300+
3x
Rather than assume what users needed, I led qualitative research and usability studies to understand the core reasons behind AI hesitation.
Findings:
• Users didn’t trust AI to make critical decisions—they needed transparency.
• They feared losing control—automation needed to be an assistive tool, not a replacement.
• Confusing UX patterns made AI interactions feel opaque rather than intuitive.
• Users wanted proof AI worked before fully integrating it into their workflows.
With these insights, I led the design of an AI experience that prioritized control, clarity, and gradual adoption.
• Explainability – Clear, in-context explanations of what the AI was doing and why.
• Human-in-the-loop – Users could override or adjust AI-generated recommendations, ensuring control.
• Small, high-impact wins first – Instead of pushing full automation, we introduced AI in low-risk, high-impact areas to build trust.
• Familiar UX patterns – AI interactions were embedded within existing workflows, making it feel like an enhancement rather than a disruption.
Iteration 1: Text-only chatbot using a partner's website
AI summary between AI agents and prospects
To future-proof AI adoption, I:
• Built a scalable AI interaction framework for consistency.
• Developed best practices for balancing automation and human input.
• Established feedback loops to continuously refine AI responses.
The redesigned experience not only improved user adoption but also contributed directly to business objectives:
• 65% AI agent adoption – Increased user trust and reliance on automation.
• 60% faster task completion – AI reduced manual workload.
• 50% fewer errors – Improved accuracy in AI-assisted tasks.
“I love how the AI saves me 5hours a day to get people truly interested in leases!”
“Very pleased with the platform. It seems easier compared to other ILSes and runs much faster.”
This project reinforced a critical lesson: AI isn’t just about capability—it’s about credibility. Even the most powerful automation won’t succeed if users don’t trust it.
• User trust must be earned – Transparency and control drive adoption.
• AI should feel like an enhancement, not a disruption – Meet users where they are.
• Start small, prove value, then scale – Early wins build confidence in AI solutions.By shifting the focus from “Look at what AI can do” to “Here’s how AI helps you”, we turned skepticism into engagement.
