Palantir

What distinguishes an AI agent from a plain LLM call?

Describe the typical agent execution loop.

How does least privilege tool scoping reduce security risk in an agent system?

Where sparse (BM25/SPLADE) retrieval reliably beats dense embeddings

Compare CrewAI, AutoGen, and LangGraph coordination models and their reliability tradeoffs

List SWE-bench, AgentBench, GAIA, and TAU-bench: what they measure, their shared blind spots, and how to interpret agent benchmark numbers.

Which practices reflect the principle of least privilege applied to an agent's tool set?

What LangGraph feature allows the agent loop to cycle back to an earlier node rather than terminating?

Prompt caching shows up on Anthropic and OpenAI pricing pages, what does it do?

When is a fixed LCEL chain preferable to an agent loop, and what is the decisive question to ask?