Modernizing data collection at FINRA with Amazon DocumentDB
summary
FINRA modernizes a high‑throughput intake pipeline that handles up to 550 billion market events per day, 110 PB of streaming storage, and spikes to 150,000 processing nodes by moving filing storage and change events into Amazon DocumentDB. It implements N+1 versioning (2.5M current documents, 12M history documents, ~5 versions average), custom field-level encryption with AWS Encryption SDK and KMS, publishes change streams for downstream workflows, and shrinks costs and latency (over 50% cost saving with Graviton); a new form ships in four weeks and customer processing time drops ~99%.
AI-written summary · about
chapters (8)
- 0:00Intro
- 3:19FINRA data AWS system
- 6:29Broker dealer registration forms
- 9:16FINRA data collection requirements
- 15:05DeCaF-FINRA next data collection framework
- 20:34Data collection statistics
- 26:49Business benefits
- 29:03Domain services-Reference architecture
related talks
38mFINRA and Fannie Mae deliver data resilience with AWS
46mNoSQL, Data Processing, and Business Transparency with RavenDB
53mManaging enterprise JSON workloads with Amazon DocumentDB
32mFINRA's knowledge graphs improve regulatory investigations
59mModernize your applications with purpose-built databases
43mFuture-proofing your applications with AWS databases
32mFINRA's knowledge graphs improve regulatory investigations
29mDelivering Always-On Financial Services with