Datarize
Frontend Engineer
Jan 2026 - Apr 2026
B2B SaaS platform for CRM marketing automation based on e-commerce customer data.
Improved widget-creation completion with AI prompt-based widget design
- Identified high abandonment in the manual widget-design setup flow
- Designed and implemented an AI chat feature that generates widget designs from natural-language prompts
- Improved widget-creation completion from 11% to 37% by refining the design flow
Reduced client engineering dependency and support requests with easy widget insertion
- Identified recurring client engineering support and CS needs caused by script-based installation
- Implemented logic that extracts the selected DOM element path and inserts widgets dynamically at that location
- Enabled marketers to install widgets directly without engineering resources
Improved recommendation accuracy and exposure efficiency through Cafe24 integration
- Identified reduced recommendation efficiency when items already in the cart reappeared in recommendation widgets
- Integrated the Cafe24 Front API SDK and cart lookup to exclude cart items from recommendations
- Skipped unnecessary API calls in test mode to reduce request costs
Reduced maintenance cost and strengthened UI consistency through design-system migration
- Identified reduced consistency and maintainability from mixing a legacy design library with a new design system
- Built a gradual release, Playwright visual-regression, and Storybook-based verification process
- Documented component mapping rules and verification steps into a repeatable migration guide