AI Product Creation Workflow
FB Catalog
Product Design | 2026
Background
As AI capabilities continued to evolve, our team began exploring new workflows that could improve efficiency and accelerate product delivery. Rather than integrating AI into a single stage of the design process, we wanted to test its impact across the entire product lifecycle. To validate this approach, I led an experiment focused on creating a new product experience from scratch while leveraging AI at every stage- from research and product definition to design and front-end development. The goal was not only to launch a working product, but also to evaluate how AI could reshape the way we build products as a team.
The Mission
Work Process
For this experiment, we selected a product page designed for users arriving directly from Facebook catalog campaigns. Since Facebook initiatives move quickly and require rapid iteration, this type of project provided the ideal environment for testing a new AI-powered workflow with the potential to scale across future projects.
01 Research
Market Understanding
As part of a separate initiative, I had previously created an AI-powered research engine using specialized agents. For this project, I used that system to define the product strategy and identify the most valuable opportunities.
The research process helped uncover which features should be included, what user expectations existed around the experience, key pain points that needed to be addressed, and areas where we could create additional value. Instead of manually collecting information from multiple sources, the research engine accelerated the discovery process and helped establish a stronger product foundation.
02 Product
Product Definition & Content Creation
Once the research phase was complete, I finalized the product requirements and began preparing all of the assets needed for production.
This included creating image prompts, content prompts, messaging frameworks, and structured content guidelines tailored specifically to the product experience. The goal was to ensure that every element generated later in the process aligned with the product strategy and user needs identified during research.
03 UX Decisions
Structure & Wireframing
I began by creating an initial concept and layout structure that could serve as a foundation for exploration inside Figma Make. Using AI-assisted generation, I rapidly tested different approaches for component placement, hierarchy, content structure, and page composition. Once I identified the direction that best matched the product vision, I transferred the selected concepts into Figma and refined them further.
04 UI Design
Visual Design
The design workflow was then connected to Cursor through MCP integrations, allowing me to continue building and iterating using AI-assisted tools. With the help of custom skills and workflows installed within Cursor, I transformed the initial concepts into a polished final design. AI accelerated repetitive tasks and exploration, while design decisions, prioritization, and refinement remained guided by product and UX considerations.
05 Development
Front-End Development
Once the design was finalized, I moved directly into implementation.
Using Cursor, I built the experience inside the company’s existing repository while following the established front-end architecture and development conventions. This allowed me to validate design decisions in a real environment and iterate on both the visual experience and product behavior until the result matched the intended user experience.
Final Result
After completing the implementation, the project was handed off to the development team through a pull request. The team then connected the front-end experience to the Facebook infrastructure and backend systems.
The experiment demonstrated the potential of an AI-assisted workflow across the entire product lifecycle. By combining AI tools with existing design and development processes, we reduced production time by approximately 60%, delivered a production-ready experience with minimal team involvement, and established a scalable framework for future projects.
For more work
noyadani3010@gmail.com