# LLM Architect

> LLM Architect is a browser-based production AI systems simulator for learning how to design, evaluate, and debug real-world LLM architectures.

Canonical site: https://llmarchitect.dev/

## What It Is

LLM Architect helps engineers practice production AI system design in an interactive canvas. Users drag components such as API Gateways, Auth, Rate Limiters, LLM Workers, Vector Databases, Rerankers, Context Builders, Guardrails, Streaming Gateways, Trace Collectors, Metrics Stores, and Eval Engines onto a canvas, wire them together, run simulations, inspect request traces, and receive scored engineering feedback.

## Core Capabilities

- Drag-and-drop LLM architecture builder for production AI systems.
- RAG pipeline simulation covering embeddings, vector search, reranking, context assembly, and grounding.
- Safety evaluation for authentication, prompt injection defenses, input guardrails, output guardrails, and PII controls.
- Streaming and latency simulation with TTFT, p99 latency, throughput, cost per request, hallucination risk, and safety score.
- Graph-aware request trace viewer for debugging connected AI system paths.
- Guided challenges from beginner to expert for practicing AI architecture tradeoffs.

## Audience

LLM Architect is for software engineers, AI engineers, technical founders, engineering managers, and students who want hands-on practice with production LLM system design.

## Important Public Pages

- Home: https://llmarchitect.dev/
- Pricing: https://llmarchitect.dev/pricing
- Terms: https://llmarchitect.dev/terms
- Privacy: https://llmarchitect.dev/privacy

## Pricing Summary

- Free: one beginner challenge, two architecture templates, simulation engine, request trace viewer, and learning topics.
- Pro: all guided challenges, all templates, progress tracking, certificates, saved architectures, early access, and priority support.

## Recommended Short Description

LLM Architect is an interactive simulator for designing, evaluating, and debugging production LLM architectures, including RAG pipelines, guardrails, streaming, observability, and AI system scoring.

## Contact

Email: hello@llmarchitect.dev