Faster, Cheaper, Greener: Pick Three
summary
FundApps rearchitects a single-process Windows monolith into a cloud-native, per-client AWS system using Step Functions, SQS, Fargate/ECS, S3, Dynamo DB and Aurora serverless V2 to eliminate noisy neighbors and always-on servers. It handles tens of millions of positions, 100 jurisdictions and 500 rules for 130 clients (~15% of global AuM, ~$17T). Moving large JSON blobs to S3 and distributing compute lets the team migrate 90% of clients migrated and fixes the last 10% with Aurora serverless V2, cutting runtimes (1h40 → 39m) and energy waste.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:38UPS mission and context
- 11:43Monolithic rule engine challenges
- 17:56Rebuilding engine in cloud
- 21:12AWS architecture overview
- 29:43Impact and client success
- 36:38Sustainability outcome
- 43:10Greener software steps
- 47:36Call to action
related talks
57mFaster, Cheaper, Greener: Pick Three
46mArchitecting for Success and Runway
46mHighly Scalable Image Storage with AWS Serverless
25mBest practices for advertising and marketing technology on AWS
10mEnergy-Efficient and Cost-Optimized Serverless Computing
29mEvolve your web application for scale
37mReactive Architecture on AWS: from 0 to 1.6M events/minute
20mSoftware Performance Optimization and Scaling