Case Study: Appsee achieves scalable real-time mobile analytics and session replay with MongoDB

Appsee Shapes the Mobile Revolution with Real-Time Analytics Service Powered by MongoDB

Appsee is a real-time mobile analytics platform that gives product teams “qualitative analytics” — session replay, touch heatmaps, user flows and crash analytics — so they can see not just numbers but why users behave the way they do. Faced with a firehose of high‑velocity time‑series session data from large customers (eBay, Samsung, Virgin, etc.), Appsee needed a database that could ingest billions of events, scale elastically, support low‑latency reads for analytics, persist complex rapidly changing data, and let a small engineering team move fast.

Appsee models each complete user session as a single MongoDB document and uses MongoDB’s aggregation pipeline, secondary indexes and a 5‑shard cluster (3‑node replica sets per shard on AWS) to ingest, analyze and serve real‑time insights. The deployment stores ~20 billion documents (15 TB) and handles ~50k ops/sec with a 50:50 read/insert split, supports predictable 2x growth, and benefits from faster developer onboarding, simplified architecture, high uptime and a 30% storage reduction after moving to WiredTiger — enabling faster time‑to‑market and reliable delivery of deep mobile analytics.

AppSee

Yoni Douek

Chief Technology Officer