Fighting crime and protecting vulnerable people
summary
A lightweight transaction-monitoring tool called ERMI processes CSV bank files, applies concrete Gherkin rules (eg flagged frequent small payments tied to Care Quality Commission records), and emits Excel reports rather than a heavy UI. Implementation uses JavaScript workers in Docker, run via AWS Lambda/AWS Batch for parallelism (eg 26-way sharding), tests from the second commit, and a deliberate choice to avoid standing infra. Result: cheap, auditable screening that already runs for large customers and reduces attack surface.
AI-written summary · about
chapters (6)
- 0:00Introduction and team
- 2:29What Ermi does
- 7:01Collaboration challenges
- 13:37Building a lightweight product
- 22:02Technical approach
- 25:53Q&A and summary
related talks
45mAI Tools in Regulated Banking: Monzo's Control Model
21mHyperautomating Compliance and Risk Intelligence in Finance
35mSecuring digital services
40mFighting Financial Crimes as an SRE
05mPutting an {ellmer} AI in production with the blessing of IT
37mTaking Elixir Live at a Bank
22mSecurity vs Dev Experience Threat … or Opportunity
32mCatch Me If You can : Continuous Delivery vs. Security Assurance