# AI use cases for AI System Design (amitshekhariitbhu/ai-system-design)

Repo: [amitshekhariitbhu/ai-system-design](https://github.com/amitshekhariitbhu/ai-system-design) | Use cases: 4 | Page: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/

A Markdown guide that teaches AI system design step by step, covering LLM inference, RAG, AI agents, MCP, guardrails, evaluation, and interview preparation.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Prepare for an AI system design interview: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/prepare-for-ai-system-design-interview/
- Plan a RAG architecture for document Q&A: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/plan-rag-document-qa-architecture/
- Design an AI agent with tools and memory: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/design-ai-agent-with-tools-and-memory/
- Estimate tokens, GPUs, and cost for an LLM feature: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/estimate-llm-feature-cost-and-gpus/

## How to prepare for an AI system design interview with AI System Design

Repo: [amitshekhariitbhu/ai-system-design](https://github.com/amitshekhariitbhu/ai-system-design) | Difficulty: Beginner | Time: 60 minutes | Basis: From the README

Use the guide's framework and interview questions to practice designing AI systems.

**Who it is for:** A software engineer with an upcoming AI or LLM engineer interview that includes a system design round.

**The problem:** Interviewers ask open-ended AI system design questions, and the candidate needs a repeatable method and vocabulary.

**The outcome:** The candidate can walk through requirements, architecture, evaluation, and monitoring for an AI system in a structured way.

### Steps

1. Open https://github.com/amitshekhariitbhu/ai-system-design and read the AI System Design Learning Path section.
2. Read How to Solve Any AI System Design Problem, Steps 1 to 8, and note the order of the framework.
3. Pick one practice problem and write short answers for Requirements, AI Objective, and Architecture Design.
4. Work through the AI System Design Interview Questions section and answer each one out loud in under five minutes.
5. Check your terms against the AI System Design Key Concepts Glossary and fix any gaps.

### Prompt to paste into your agent

```text
Using the repo amitshekhariitbhu/ai-system-design, read the section How to Solve Any AI System Design Problem and apply its eight steps to the practice problem <target>. Output a one-page design outline with one heading per step. Verify that the Evaluation and Monitoring steps are both present.
```

### Caveat

The repo is a written guide with no code or runnable setup, so it teaches design thinking rather than producing a working system.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/prepare-for-ai-system-design-interview/

## How to plan a RAG architecture for document Q&A with AI System Design

Repo: [amitshekhariitbhu/ai-system-design](https://github.com/amitshekhariitbhu/ai-system-design) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README

Use the RAG sections to choose chunking, search, and reranking for a document assistant.

**Who it is for:** An AI engineer who must build an assistant that answers questions over company documents and needs to choose the retrieval design first.

**The problem:** Retrieval quality decides whether a RAG assistant answers correctly, and the design choices for ingestion, chunking, and search are easy to get wrong.

**The outcome:** A written RAG design that names the parsing, chunking, hybrid search, and reranking choices with reasons.

### Steps

1. Open the RAG (Retrieval Augmented Generation) section of the README at https://github.com/amitshekhariitbhu/ai-system-design.
2. Read Document Parsing and Ingestion and Chunking Strategies, then write down the chunking approach for your documents.
3. Read Hybrid Search and Reranking, then decide whether your queries need keyword matching, semantic matching, or both.
4. Read the Vector Database section and pick an index type for your data size.
5. Write a one-page design that lists each choice, the reason for it, and how you would measure retrieval quality.

### Prompt to paste into your agent

```text
Using the repo amitshekhariitbhu/ai-system-design, read the RAG section and produce a retrieval design for <target>, covering chunking, hybrid search, and reranking. Output the design as a Markdown document with one section per choice. Verify that each choice cites the README section it came from.
```

### Caveat

The README explains the RAG options but does not provide code, so you still need to implement and test the chosen design yourself.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/plan-rag-document-qa-architecture/

## How to design an AI agent with tools and memory with AI System Design

Repo: [amitshekhariitbhu/ai-system-design](https://github.com/amitshekhariitbhu/ai-system-design) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README

Use the agent, tool calling, MCP, and memory sections to scope an agent's design.

**Who it is for:** An agent engineer who has been asked to build an AI agent that calls internal tools and remembers past sessions.

**The problem:** Agents fail in predictable ways, and it is hard to decide how tools, MCP servers, and memory should fit together.

**The outcome:** A design for an agent that lists its tools, MCP usage, memory layers, and known failure modes to test.

### Steps

1. Open the AI Agents and Agentic Systems section and read The Five Core Parts and Common Failure Modes.
2. Read Tool Calling and Model Context Protocol (MCP), including When to use MCP and When MCP is overkill.
3. Read Memory for AI Agents and decide which memory operations and layers your agent needs.
4. Read Multi-Agent Systems only if your task needs more than one agent, and note the trade-offs section.
5. Write the design as a list of tools, memory stores, and failure cases with one test for each failure case.

### Prompt to paste into your agent

```text
Using the repo amitshekhariitbhu/ai-system-design, read the AI Agents, Tool Calling, MCP, and Memory sections and draft a design for the agent <target>. Output a table of tools, memory layers, and failure modes, each with a test. Verify that MCP is recommended only if the README's When to use MCP criteria are met.
```

### Caveat

The guide describes agent design concepts and does not ship an agent framework, so the implementation depends on tools you choose separately.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/design-ai-agent-with-tools-and-memory/

## How to estimate tokens, GPUs, and cost for an LLM feature with AI System Design

Repo: [amitshekhariitbhu/ai-system-design](https://github.com/amitshekhariitbhu/ai-system-design) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README

Use the estimation and cost sections to size tokens, GPUs, and spend before launch.

**Who it is for:** A platform or ML engineer who must give a budget and hardware estimate for a new LLM feature before it is approved.

**The problem:** Teams often launch LLM features without estimating token volume, GPU needs, or cost per request, then face surprise bills.

**The outcome:** A rough estimate of tokens per day, GPUs needed, and monthly cost, with the assumptions written down.

### Steps

1. Open the Back-of-the-envelope Estimation for AI section and read Token Estimation, GPU Estimation, and Cost Estimation.
2. Read Tokens, Latency, and Throughput to understand TTFT, TPS, throughput, and cost per token.
3. Write down your expected requests per day and average input and output tokens per request.
4. Apply the formulas from the estimation sections to get tokens per day and GPUs needed.
5. Read Cost Optimization and list which of the seven options you will test first.

### Prompt to paste into your agent

```text
Using the repo amitshekhariitbhu/ai-system-design, follow the Back-of-the-envelope Estimation for AI section to estimate tokens, GPUs, and cost for <target>. Output a short table of the inputs, formulas, and results. Verify that every formula used appears in the README and that each assumption is stated.
```

### Caveat

The estimates depend on your own traffic and model numbers, which the README does not supply, so treat the output as a starting point rather than a quote.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/amitshekhariitbhu-ai-system-design/estimate-llm-feature-cost-and-gpus/