← All discussions

Finance & Operations

Risk Committee Pushback on New AI‑Driven Stress Testing Framework

Krishna Robinson · 19 Oct 2026

Risk officers are being asked to embed the Bank of England's new AI‑driven stress testing model into the quarterly reporting cycle. We've discovered the model needs daily data refreshes and real‑time scenario generation, yet our legacy Basel‑III platform only supports monthly uploads, creating a compliance bottleneck. What practical work‑arounds have other firms used to reconcile the AI model's frequency demands with existing reporting systems without a full‑scale rebuild?
👍 401 💬 2

2 replies

R

Richard Johnson · 20 Oct 2026

We tried to push daily AI outputs into our old Basel‑III system with a quick Python script, but it blew up because the platform only accepts monthly timestamps.
We've since set up an Azure Data Factory job to roll the daily results into a monthly CSV and drop it into the legacy upload, which buys us a few weeks while we press our vendor FIS for an API fix.
A

Arianna Bishop · 20 Oct 2026

I once set the AI feed to run at midnight and crashed the core server. I couldn't have known better than to use a simple Azure Data Factory job that writes to a Snowflake landing zone, which would have let us keep the Basel‑III uploads on their schedule while the model ran on fresh data.

Log in or create a free account to join the conversation.