ReAct Loop Animation
Prompt chaining, routing, orchestrator-workers, ReAct, plan-and-execute, reflection and tree search: what each costs, what the research shows, and when to climb to the next one.
Animate: ReAct Loop Animation
Step through the flow one stage at a time and watch the data move.
Reading Agent Design Patterns is free. Pro unlocks the other four modes on all AI Engineering topics, so you can go from understanding it to building it.
Helps to know first
- SeeThe architecture as diagrams, so you can hold the whole system in your head.
- AnimateStep through the flow one stage at a time and watch the data move.
- TestScenario questions with explanations that show exactly where your model of it breaks.
- BuildA guided implementation with code you can run and adapt for production.
Continue reading
All AI Engineering articles →Model Serving With vLLM: The KV Cache Decides Your Capacity
KV cache arithmetic, PagedAttention, continuous batching, prefix caching, FP8 and speculative decoding, the vLLM flags that matter, and the five metrics that show which limit you hit.
Voice Agents: A Real-Time Problem With a Model Inside
Speech-to-speech or a pipeline, a latency budget for every stage, interruptions that really work, prompts written for the ear, and what a call actually costs.
Guardrails: Input/Output Safety in Production
Input, output, schema, tool, and cost guardrails — the layered defence system that catches LLM failures before they reach a user or downstream system.
Discussion
Sign in to share your feedback and join the discussion.