Samsung Moves to Real-Time Pricing with AWS Lambda Streaming

Samsung's product listing pages needed to display pricing for dozens of SKU combinations during high-traffic events, and the legacy Data Aggregation (DA) layer handled this via an hourly cron job that precomputed every possible product permutation into a cache. This approach was both storage-heavy—since most precomputed combinations were never actually requested—and slow to reflect price changes, sometimes lagging up to an hour behind the pricing engine. During flash sales this meant customers saw stale prices, and checkout mismatches eroded trust.

Working with AWS, Samsung engineers replaced the intermediate cache with a stateless Bulk Arbitration Engine that queries the pricing engine directly at request time. Using AWS Lambda Response Streaming, the system fans out parallel requests and streams results back to the client as NDJSON without buffering, letting the browser render prices as they arrive rather than waiting for the full payload.

After evaluating buffered Lambda invocations and improved EC2 caching, the team chose Response Streaming because it was the only option enabling parallel fan-out, low time-to-first-byte, and a fully stateless design with no cache to invalidate. CloudFront absorbs most edge traffic, reducing direct load on Lambda. The pattern offers a practical blueprint for engineers building high-traffic APIs that aggregate data from multiple backend sources.