Lyft Case Study

Lyft is a San Francisco-based ridesharing company that is on Fortune magazine’s “Unicorn” list of hot startups, with a valuation of $5.5 billion. Founded in 2012 and with 315,000 active drivers in more than 190 U.S. cities, Lyft is the fastest growing company of its kind in the United States, reporting 200 percent year-over-year growth in 2015 for both rides and revenue. Lyft, with a mission of reconnecting people and communities through better transportation, is working with General Motors to create a network of on-demand autonomous vehicles and has partnered with local market leaders Didi Kuaidi, Ola, and GrabTaxi to launch its international expansion in Asia.

• Runs a microservices infrastructure on AWS.
• Uses Jenkins orchestration of a Continuous Integration (CI) pipeline; Jenkins manages the builds and tests, automating the process.
• Deploying code into production involves a series of tests that require large pools of compute resources, which led to high costs for routine daily tasks

• The Infrastructure team found it could save up to 90 percent on its CI processes by using Amazon EC2 Spot instances with more compute capacity instead of using higher-priced on-demand instances.
• Used four lines of code to modify a Salt module that launches new instances from the Spot market.
• CI processes do not require a lot of power for testing, enabling use of older-generation, lower-cost Amazon EC2 instances.

• Saves up to 75 percent monthly for testing processes compared to using on-demand instances.
• Can leverage older-generation instances with greater memory resources to run more containers, allowing tests to finish faster.