BREEZY AI
PROJECT NAME
BREEZY AI WORKFLOWS
YEAR
2026
ROLE
AI PRODUCT DESIGNER
Overview
BreezyAI is an AI native startup based in Silicon Valley offering a B2B SAAS AI-powered front office platform to capture leads, answer calls and messages 24/7, automate follow-ups, manage CRM tasks.
TL;DR
We completed a redesign of the Inbound AI experience and introduced a new workflow feature that allows users to harness the AI agent when handling incoming calls. We then added more features to allow more user control and freedom.
Project Overview
Business Problem
Solopreneurs can’t answer every call on the job.
Missed calls result in missed revenue.
Setup for AI workflows often feels technical or opaque.
Success Metrics
Activation & Lead Conversion
Automation Adoption & Usage
Ongoing Value & Stickiness
Design Process
AI-First Design Approach
My core design philosophy has always been to equip users with the power of AI, rather than replace them. At Breezy AI, this meant redesigning our AI receptionist experience so small businesses could shift their focus from answering phone calls to executing their core craft. Working within a lean, AI-native team powered by early iterations of Claude Code, traditional roles converged. This dynamic environment anchored my design process: I focused on building intuitive, bespoke solutions that allowed users to seamlessly harness their AI agents within existing workflows, requiring a continuous, conscientious balance of speed, strategic direction, and design craft.
Design Process
From AI Slop to Levels of Determinism
When I joined as the founding designer, the team—comprising just one engineer and one product manager—was already shipping rapidly using Claude Code. While their velocity was impressive, it created a sort of product tunnel vision. To bridge the gap between rapid AI generation and intentional design, I utilized Figma MCP to bring their code directly into Figma. Once we established this bidirectional workflow, my focus shifted toward aligning our output on structure, context, and user behavior. This holistic approach drove our complete redesign of the AI receptionist, enabled new workflow capabilities, and allowed us to refine existing features—like Notifications—to match users' actual mental models.
Design Process
Fine-Tuning the Final 10%
The true challenge of AI-native product design lives in the last 10% of the prototype. As we rolled out new capabilities, we relentlessly evaluated their inherent value—asking if this truly saves a busy solopreneur time on the job. Thanks to our rapid workflow, I moved away from pitching a single design approach. Instead, I focused our team discussions on exploring multiple dimensions of a problem—fluidly pivoting, combining ideas, and refining the details until we landed on the optimal solution to ship for real-world feedback.
Key Results
Impact & Outcomes
In just four months, our AI-accelerated workflow enabled us to launch a major end-to-end redesign, optimize the core onboarding flow, and ship five net-new features and enhancements. The impact of this rapid execution culminated in my client receiving multiple acquisition offers by the close of our engagement. More than just a successful product launch, this project was a catalyst for my own evolution, permanently embedding AI tools into my design process.
Curious how I would tackle this same challenge using today’s ever-evolving AI capabilities? Reach out, and let’s talk.
