morganmcg1

sean goedecke
Everything I know about good system design
I see a lot of bad system design advice. One classic is the LinkedIn-optimized “bet you never heard of queues” style of post, presumably aimed at people who are new to the industry. Another is the Twitter-optimized “you’re a terrible engineer if you ever store booleans in a database” clever trick1. Even good system design advice can be kind of bad. I love Designing Data-Intensive Applications, but I don’t think it’s particularly useful for most system design problems engineers will run into.

What is system design? In my view, if software design is how you assemble lines of code, system design is how you assemble services. The primitives of software design are variables, functions, classes, and so on. The primitives of system design are app servers, databases, caches, queues, event buses, proxies, and so on.

This post is my attempt to write down, in broad strokes, everything I know about good system design. A lot of the concrete judgment calls do come 

morganmcg1

import os
import weave
from langchain_google_genai import ChatGoogleGenerativeAI

os.environ["GOOGLE_API_KEY"] = 

GEMINI_MODEL_NAME = 'gemini-2.5-flash'
CUSTOM_MODEL_NAME = 'lc_' + GEMINI_MODEL_NAME 

weave_client = weave.init("cost-testing")

morganmcg1

# uv pip install weave langchain-google-genai -qq

import os
import weave
from langchain_google_genai import ChatGoogleGenerativeAI

os.environ["GOOGLE_API_KEY"] = 

GEMINI_MODEL_NAME = 'gemini-2.5-flash'
CUSTOM_MODEL_NAME = 'lc_' + GEMINI_MODEL_NAME 

morganmcg1

[
  {
    "index": 0,
    "question": "Hey I have a question about using wandb with fastapi in a prod environment. is it recommended to initialize wandb within a specific route function, ie\n\n    `@app.route('/')\n    def my_function():\n        wandb.init(...)\n    `\n    or should i initialize beforehand:\n\n    `wandb.init(...)\n    @app.route('/')\n    def my_function():\n        ...`\n\n    I'm getting a long list of log items in the console and many of them are empty.",
    "generated_answer": "We recommend initializing wandb as early as possible in your FastAPI application, preferably before defining your route functions. This helps capture any console output, including error messages, which can be useful for debugging.\nTo set up wandb in your FastAPI app:\nInitialize wandb at the start of your main application file:\nimport wandb\nwandb.init(project=\"your_project_name\")\n\napp = FastAPI()\n\[email protected]('/')\ndef my_function():\n    # Your route logic here\n\nTo address the empty log items, try inc

morganmcg1

[
    {
        "index": 0,
        "question": """Hey I have a question about using wandb with fastapi in a prod environment. is it recommended to initialize wandb within a specific route function, ie

    `@app.route('/')
    def my_function():
        wandb.init(...)
    `
    or should i initialize beforehand:

morganmcg1

Below is a practical “cheat‑sheet” distilled from the many production prompts you just saw (Cursor, v0, Devin, Manus, Replit, Lovable, etc.). Follow these guidelines when designing prompts for your own LLM‑powered apps and you’ll get more reliable, safer and higher‑quality results.

────────────────────────────────────────────────────────

1. Nail the Three Core Elements
   ──────────────────────────────────────────────────────── A good prompt almost always contains ❶ role, ❷ objective, ❸ constraints / format.

• Role
“You are a senior Rust compiler engineer…”
Sets the persona, vocabulary and depth.

morganmcg1

ources.
Browsing completed
Browsing completed
I am browsing the URL: https://llmstxt.org/
Browsing completed
I am browsing the URL: https://weave-docs.wandb.ai/
Now that I have a good understanding of both the llms.txt concept and Weave's documentation, I'll create an appropriate llms.txt file for Weave. I'll focus on making it useful for LLMs to understand and work with Weave's capabilities.
Edited the contents of a file
I've created a comprehensive llms.txt file for Weave that follows the standard format and provides structured information about the framework. Here's what I included:
Title (H1): Clear identification of the project

morganmcg1

# Weights & Biases (W&B) Weave

> Weave is a framework for tracking, experimenting with, evaluating, deploying, and improving LLM-based applications. It provides comprehensive tools for tracing LLM calls, monitoring application behavior, systematic prompt engineering, evaluation, and deployment of guardrails in production.

Weave is designed to support the entire lifecycle of LLM application development, from initial experimentation to production deployment. Key capabilities include:

- Tracing and monitoring of LLM interactions and application logic
- Systematic iteration on prompts, datasets, and models
- Experimentation through an LLM Playground
- Comprehensive evaluation tools with custom and pre-built scorers

morganmcg1

from weave.scorers import WeaveToxicityScorerV1

toxicity_scorer = WeaveToxicityScorerV1()


@weave.op
def call_llm(prompt: str) -> str:
    """Generate text using an LLM."""
    # Your LLM generation logic here
    return prompt.upper()

morganmcg1

<file_tree>
/Users/morganmcguire/ML/marimo
├── .github
├── ├── ISSUE_TEMPLATE
├── └── workflows
├── configs
├── dagger
├── └── src
├── └── └── main
├── development_docs