DataFacts: AEP’s first Data Anomaly Detection Tool built on App Builder
summary
Softcrylic builds DataFacts on App Builder (formerly Project Firefly) and combines schema profiling with Great Expectations rules plus active‑learning ML to flag anomalies, completeness drops, validity failures and dupes. It profiles three months on install, sets a default 5% threshold, then scans each new batch and sends in‑app/email alerts and SQL exports. Demos show it flagging 18 million records with 20,000 failed checks, detecting 4,000,000 vs 200,000 missing‑data spikes and a 28‑profile cluster of duplicates, with dashboards and drilldowns for root‑cause.
AI-written summary · about
chapters (16)
- 0:00Intro
- 1:25Why DataFacts
- 3:52Approach
- 6:23Features
- 8:33Demo
- 9:40Data Quality
- 11:40Data Quality Reports
- 12:13InApp Notifications
- 13:23Filters
- 15:08Error Validation
- 15:52Reporting
- 17:55Validation
- 20:15Data Quality Validation
- 23:25Duplicate Customer Profiles
- 24:57Customer Profiles from AEP
- 26:40Summary
related talks
01h 00mData is out, knowing is in: Applying AI to automate cloud operations
14mAccelerate Scalable, Secure Apps with APEX and GenAI
28mHow to boost data science routines with DataArt’s AutoML Platform?
48mUnlocking the Future: See How AI Is Transforming Analytic Applications
28mNew normal or not normal? Observability that will change the way you work
11mBig Data Moonshots and Ground Control
01h 02mOracle Fusion Cloud Applications—Secure, Extensible
30mDetection Allegro: Composing Detection Rules with... (Raphael Ruban, Chen Cao)