AAPICODE.IO
Projects & Case StudiesAdvanced 7 min readUpdated 2026-04

Case Study — Decisioning Platform at Jenius Bank

How the Credit, Fraud & Risk Technology team rebuilt the decisioning platform: cut P99 from 200ms to 5ms, scaled throughput 3.3x, and shipped 96+ releases without P1 incidents.

Risk TechDecisioningPerformanceMigration

Context

Founding the Credit, Fraud & Risk Technology (CFRT) team at Jenius Bank meant standing up an end-to-end personal loan product on top of integrated bureaus, aggregator channels, and a modernized rule-engine stack.

What we shipped

  • Lending Tree integration unlocking $30M+ in monthly origination and 30,000+ leads/month.
  • Migration from a legacy SaaS rule engine, cutting P99 latency from 200ms to 5ms.
  • Throughput scaling from 72 to 240 RPS for core decisioning (3.3x growth).
  • Pricing deployment cycle compressed from 3 weeks to 1 week.
  • 96+ production releases in FY24 with zero P1/P2 incidents and 60–80% more automated tests.

How we did it

  • Configuration-driven orchestration with versioned strategies for rule sets and integrations.
  • Bureau integrations with TransUnion, Equifax, and LexisNexis for layered fraud signals.
  • Rule-engine modernization on Provenir, IBM-ODM, and Drools with strong test harnesses.
  • Release automation, golden-path CI/CD, and per-strategy observability.