The team needed earlier visibility into which accounts were most likely to churn instead of reacting after revenue risk had already surfaced.

Built a churn analysis workflow using behavioral, usage, and account level signals to identify patterns tied to early drop off and declining engagement.

Created a clearer view of at risk accounts and helped frame retention decisions around measurable product behavior.

Customers were dropping out across the checkout journey, but the biggest sources of friction were not clearly understood.

Mapped the funnel across devices, user segments, and journey stages to find where conversion loss was highest and which user groups were most affected.

Turned a broad conversion problem into specific areas of focus for checkout, messaging, and UX improvement.

The challenge was understanding which lapsed users were most likely to return and what behavioral signals mattered most.

Analyzed churned user behavior and built segments around listening patterns, inactivity windows, and re engagement triggers.

Helped structure retention thinking around user behavior instead of broad assumptions.

Duolingo’s growth feels simple on the surface, but its retention engine comes from tightly connected product loops and behavior design.

Broke down the product using engagement, growth, and habit formation frameworks to understand how its mechanics support repeat usage.

Created a structured product analysis showing how feature design, streak behavior, and motivation loops reinforce retention.

Semantic search alone was not enough to surface the most useful creator results for real business needs.

Built a hybrid retrieval approach that combined semantic similarity with keyword based matching to improve relevance and ranking quality.

Produced a more practical search experience that balanced contextual fit with direct business usefulness.