Guardrails or it fails
You've gotten the to point where you're pretty happy with the agent's performance and costs have been optimized. Before scaling it up to 50% of Nile's users, you first have to pass an audit. The Head of Infrastructure wants to know what guardrails you've put in place to mitigate misuse of this system.
Which of the following are legitimate guardrails to prevent misuse?
Questo esercizio fa parte del corso
Building Scalable Agentic Systems
Esercizio pratico interattivo
Passa dalla teoria alla pratica con uno dei nostri esercizi interattivi
Inizia esercizio