Building an AI experience from scratch

Company: Glassdoor

Role: Senior Product Designer

Experimental iteration: April 2025 - July 2025

MVP development: July 2025 - October 2025

A holistic design approach enables thoughtful integration across the product space. With AI implementation, conversational design requires that systemic thinking.

Experimental iteration

partnered with a cross-functional tiger team of designers, product managers, and engineers to lead a rapid experimentation program focused on validating product strategy through live testing. My role centered on translating business and user hypotheses into iterative design solutions delivered each sprint.

Experiments spanned onramp placement, messaging strategy, and UI patterns, all deployed on the production site and informed by parallel user research. Each test built on prior learnings to progressively refine the experience and guide longer-term product direction. I also conducted a competitive audit of AI surfacing and coaching experiences to identify best practices, differentiation opportunities, and strategic gaps.

I collaborated closely with a small tiger team comprised of designers, product managers, and engineering. My track focused on rapid iteration on product to test questions posed, delivering a new design per sprint.

Testings included placement of an onramp, content of the on ramp, and UI. These were all tested on our live site. This fast-paced work took into account UXR that was conducted in parallel and built on learnings from each test. Additionally, I worked to conduct a competitive audit of AI surfacing and coach.

MVP development

Building on insights from the experimentation phase, I led the design of core patterns for integrating AI into a legacy product experience, establishing the foundation for the MVP.

I partnered with the tiger team to define UI patterns for surfacing AI-generated summaries directly within existing pages, then expanded the experience into a dedicated AI surface. This included designing user input flows, prompt interactions, and AI response presentation, while collaborating closely with content strategy to define guardrails and messaging parameters. The MVP designs were delivered within two months and successfully launched in October.

Developing Global Access

Following the AI MVP launch, I expanded initial onramps into global access points through product navigation, driving broader discovery and adoption of the experience. I partnered closely with research to run rapid, iterative tests across multiple design variants, using insights to guide direction and validate key decisions.

In parallel, I collaborated with mobile engineering to define an AI-led navigation approach across native platforms. I synthesized these explorations into a strategic proposal for design leadership, weighing cross-platform consistency against platform-specific optimization to arrive at a scalable implementation strategy that prioritized user experience quality.

This explorative AI design approach, resulted in a foundational approach to non-determinative design for this product.