Learning path
Safe and Reliable AI Apps
Prompt injection defence, guardrails, moderation and privacy, and testing LLM output with evals, with animated films and runnable Python. Part of the Gen AI Developer premium track.
-
1
Prompt Injection and How to Defend Against It
The number-one security risk of LLM apps: direct and indirect prompt injection with real-looking examples, why keyword filters fail, and the layered defences that work (separate data from…
-
2
Guardrails, Moderation and Privacy
Put guardrails around an LLM app: input and output checks, keeping the bot on topic, moderation for harmful content, redacting personal data (emails, phone numbers, PAN, Aadhaar) before…
-
3
Testing and Evaluating LLM Output
Stop guessing whether a prompt change helped: build a small eval set, score answers with must and must-not checks, compare two bot versions, add an LLM judge for…