Case Study: Intercom achieves 90% cost reduction and rapid serverless scale with Amazon Web Services (AWS)

Intercom Chooses AWS Lambda to Deliver Fast Customer Value

Intercom, a messaging‑first customer engagement platform used by ~30,000 companies and powering over 500 million monthly conversations, needed to reduce operational overhead and speed innovation while protecting system availability. Engineers faced two challenges: modernizing a billing pipeline built on self‑hosted Elasticsearch and stopping customer “misbehaving” integrations that generated excessive writes and risked DynamoDB partitions.

Intercom adopted AWS serverless services (AWS Lambda, Amazon Athena, Kinesis Data Firehose and Analytics) to rebuild billing and implement a real‑time throttling system. The serverless approach cut billing costs by ~90%, saved about 800 maintenance hours per year, was prototyped and deployed in under two weeks, and scales to handle more than 1 billion updates per day; the throttling solution now detects ~50 bad connections hourly and prevents ~50,000 risky requests, with virtually no maintenance since launch.

Intercom

Rob Clancy

Engineering Manager