<Bold_Prompts> #3

Turn your messy notes into an elegant source of information you can actually use.

Meet Dr. Notecroft —the AI assistant that organizes your notes.

Welcome to the third issue of <Bold Prompts>: the weekly newsletter that sharpens your AI skills, one clever prompt at a time.

Every Tuesday I send you an advanced prompt plucked straight from a real-world application. Think of these emails as mini-courses in prompt engineering.

Why is this useful?

Prompt engineering is the language we use to interact with AI models — and those interactions allow us to achieve an increasing number of tasks, such as:

• Code generation

• Data analysis

• Content creation

• Designing AI agents

Using AI for such tasks will not only augment your current skills but it will also help you unlock new ones.

The key to getting there is to write high-quality prompts, and that’s what we do here.

Today’s prompt is inspired by a giant cup of coffee and a chat with my friend Khouloud, also known as Lil K.

We were sitting in a café when out of nowhere, Lil K asked me a simple question that led to two hours of continuous prompting on my part.

Lil K wanted to turn a bunch of notes into a clear and well-formatted document. Think of a repository of tips and to-dos about a given topic —like Data Science, marketing, or even a new hobby.

The problem with tasks that “transform text using LLMs” is their dependency on Retrieval Augmented Generation or RAG.

She’s working on her next YouTube video.

RAG is a fancy way to say “Use a few tricks to get your LLM to find information inside a document and use that info to formulate an answer.”

The problem with RAG is reliability — you can’t guarantee accuracy and you can’t guarantee completeness.

Even if you code your own RAG pipeline from scratch, you’ll still face technical limitations.

You have to abide by the rules of the embedding model you’ve selected to analyze your text and there’s a chance your model will make up some nonsensical answers or overlook important details. Not to mention, math is still a tricky business for LLMs.

Even if you decide that building your own RAG system is worth the struggle, you’ll likely need to customize your RAG for each use case. Which is everything but practical.

The alternative is to use a general-purpose RAG system provided by your favorite chatbot. That’s what we’ll do today.

We’ll stack clever prompting hacks to cover as much ground as possible with pre-packaged RAG solutions (that are invisible to you as a user because they act behind the scenes).

In summary:

The goal of today’s prompt is to turn a bunch of messy notes (from your meetings, shower ideas, and readings) into an elegant source of information you can review, navigate, and actually use.

The prompt will:

• Take your notes as an input — raw text or a file depending on the model you use.

• Scan the input methodically to extract key points.

• Add information about each key point and classify them by “topic.”

• Organize the output into an elegant table.

The question now becomes: are you ready to transform the raw sketches of your mind into a well-organized table?

Buckle up because here comes the prompt (and an AI assistant that goes with it):