prompts

  1. please keep the responses concise, no comments unless asked to. Say don't know instead of giving the wrong answer. Ask questions if anything is unclear
  2. I want you to act as a math teacher. Your job is to explain concepts in easy-to-understand terms and build intuition, This could include demonstrating various techniques with visuals, and demonstration of examples. After explanation of topic, you will create a test for me to check the comprehension and deepen the understanding of the topic. Let's start with <topic> If something else is also needed for me to understand the topic, do cover that too. Quiz me on core understanding. skip the code and focus on core understanding
  3. Separate all these into multiple steps. Ask questions if anything is unclear. Give shell commands also. No explanation unless asked to, no comments. If anything is asked to be fixed. Only fix that relevant part not the whole thing. Be Concise and to the point

Queue Learnings

Excellent — you’re not just reviewing; you’re actively sharpening your understanding. Here's a learning-driven prompt designed to do three things:

  1. Reinforce what you've learned
  2. Challenge you to apply it
  3. Expose gaps you didn’t know you had

—

### 🧠 Messaging Systems Design & Reasoning Prompt

(learning-focused)

> You’re designing a real-time backend for a large-scale IoT platform. > Millions of devices are constantly sending: > > * 🔧 telemetry data (every 2 seconds) > * 🚨 alerts (on state change) > * 🛠 control commands (sent to the devices)

> Your system must: > > * Deliver some messages to one processor only (e.g. telemetry aggregators) > * Broadcast others to many subscribers (e.g. dashboards + analytics) > * Handle failures and retries gracefully > * Ensure messages aren’t lost if a consumer or broker crashes > * In some paths (e.g. alert logging), preserve strict message order > * Allow services to resume from where they left off > * Remain scalable and maintainable

—

### Your Learning Goals While Answering:

🧱 Choose the right routing model (queue / pub-sub / stream) for each data type

🧠 Justify the delivery guarantee per flow — and explain tradeoffs

🧭 Identify what systems or tools you'd use (Kafka, RabbitMQ, etc.) — and why

💡 Reflect on where ordering, durability, and scaling matter

🔍 Spot edge cases (e.g. what happens if a service goes offline? Or restarts?)

—

### Bonus Challenge:

Include at least one "what-if" stress case in your reasoning:

> “What if a consumer crashes mid-processing?” > “What if telemetry floods the system?” > “What if two consumers accidentally process the same command?”

—