Sarah Chen

Users struggled to find relevant content across 50k+ saved items. Search satisfaction was at 32%. Leadership wanted a 2x improvement in 3 months.

Built AI-powered semantic search using Claude. Created custom embeddings pipeline, A/B tested 5 ranking algorithms, launched with progressive rollout to 10% → 50% → 100% of users.

Search satisfaction jumped to 78% (+46pts). Feature adoption hit 65% DAU within 2 weeks. Reduced support tickets by 40%. Became #1 driver of Premium upgrades.

High-value users were churning silently. No early warning system. By the time we noticed, it was too late to intervene.

Built churn prediction model using 50+ behavioral signals. Trained Claude to analyze user patterns, created automated intervention workflows for at-risk users.

Reduced churn by 28% in high-value segment. Retention team now reaches out 2 weeks before predicted churn with personalized offers. Saved $2M ARR in first quarter.

Organic growth had plateaued at 5% MoM. CAC was rising. Needed a sustainable growth channel that didn't rely on paid acquisition.

Designed a referral program with AI-generated personalized invite messages. Built incentive mechanics (both sides win), optimized for viral coefficient >1.0.

Referral signups grew from 0 to 35% of new users in 8 weeks. Viral coefficient hit 1.3. CAC dropped 60% for referred users. Program now drives 40% of growth.