← Browse

Session brief

Learned Execution Graphs for Anomaly Detection & Drift in APIs — Ritvik Pandya, JP Morgan Chase

Ritvik Pandya , JP Morgan Chase

20 min

Overview

This talk introduces Learned Execution Graphs (LEGs) as a novel approach to detecting anomalies and drift in API usage. The core thesis is that by modeling API interactions as graphs and learning their execution patterns, systems can proactively identify deviations from normal behavior, which is crucial for maintaining API reliability and security.

Who should watch

Key takeaways

Notable quotes

Ritvik Pandya discussed modeling API interactions as graphs to learn normal execution patterns.
The system identifies anomalies by detecting deviations from these learned graph structures.

Watch on YouTube →

Unofficial community note. Prefer the recording for nuance.