Full-stack engineering · AI workflows · Conversion optimization · 2026
Looplift
An AI-assisted experimentation platform connecting storefront audits, variant generation and A/B testing.
Private beta
The problem
Shopify merchants need to turn conversion findings into experiments they can review, launch and evaluate. I built Looplift around that workflow, connecting evidence collection, AI-assisted recommendations, generated variants and explicit human approval.
Engineering
The Next.js and TypeScript application uses Supabase for authentication and PostgreSQL data. An audit pipeline collects storefront evidence, analyzes pages and synthesizes findings. Variant generation produces versioned declarative patch sets, while an experiment lifecycle connects approval, delivery, measurement, guardrails and rollback.
Product decisions
Opportunities persist as distinct product records so audits and proposals can build on the same finding. Merchants review each generated change before launch. The interface brings audits, reviewed variants and experiment states together.
Current state
The project is in private beta. The audit-to-launch workflow and experiment controls are implemented. The documented evidence has not yet established a statistically conclusive experiment or a proven learning loop; those outcomes need further usage and validation.
Project visuals
Select an image to view it at full size.
