Just Five Mins!

🧠 What RAG Actually Does

RAG enhances LLMs by retrieving relevant external information (e.g. from documents or databases) at query time, then feeding that into the prompt. This allows the LLM to answer with up-to-date or domain-specific knowledge without retraining.

💸 Is RAG Expensive?

Yes, it can be — especially if:

* You repeatedly reprocess large documents for every query.

* You use high token counts to include raw content in prompts.

* You rely on real-time parsing of files (e.g. PDFs or Excel) without preprocessing.

This is where vector storage and embedding optimization come in.

📦 Role of Vector Storage

Instead of reloading and reprocessing documents every time:

* Documents are chunked into smaller segments.

* Each chunk is converted into a vector embedding.

* These embeddings are stored in a vector database (e.g. FAISS, Pinecone, Weaviate).

* At query time, the user’s question is embedded and matched against stored vectors to retrieve relevant chunks.

This avoids reprocessing the original files and drastically reduces cost and latency

⚙️ Efficiency Strategies

Here’s how to make RAG more efficient:

Strategy

Description

Benefit

Vector Storage

Store precomputed embeddings

Avoids repeated parsing and embedding

ANN Indexing

Use Approximate Nearest Neighbor search

Fast retrieval from large datasets

Quantization

Compress embeddings (e.g. float8, int8)

Reduces memory footprint with minimal accuracy loss

Dimensionality Reduction

Use PCA or UMAP to reduce vector size

Speeds up search and lowers storage cost

Contextual Compression

Filter retrieved chunks before sending to LLM

Reduces token usage and cost

Design for UI/UX is obviously an art form, but can AI do as good a job or better? or as the case may well be, using AI to help with the tedious stuff?

UX Pilot

Figma

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Okay, AI agent automation is everywhere and so are MCPs (Model Context Protocol) to allow you to connect to anything and everything via AI. However, this still needs some technical know-how, even in the low-code offerings.

But what if you have a complete virtual AI employee? and better yet, where it can work with your existing applications seamlessly - even older ERP systems or applications that have no API or external connections?

Let’s take a listen to find out more!

Warmwind OS

Warmwind OS intro

This is a deeper dive into Warmwind OS

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

So, a slightly boring episode but surprisingly useful, I think.

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Oh dear, so many meme phrases!

Hey, this is a free podcast however, if you feel you want to support me then check out Patreon. I will have some more detailed deep-dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Using VS Code (free and open source) along with the AI Toolkit extension allows you to play with several different AI models as well as local models - all from one user interface. Yes, you can run AI models locally! (okay, not the larger online ones!) and certainly the SLM (small language models) models like PHI-4 from Microsoft.

You can even get into building AI agents visually and generate actual code!

You don’t have to be a developer - this stuff is easy enough to use if you are curious about this sort of thing. If you want some guidance or examples of what you can do, just ping over a comment and I will be happy to help.

Hey, this is a free podcast however, if you feel you want to support me then check out Patreon. I will have some more detailed deep-dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

What does all this mean?

Microsoft AI Extensions: An abstract set of classes to enable a c# .Net developer to talk to different AI models/platforms instead of specific individual frameworks. Also allows you to create and interact with MCP (model context protocol).

MCP: Model Context Protocol - soooo many resources out there now. I did an episode on MCP a few podcasts ago if you want to take a look/listen.

Azure AI Foundry Local: A bit like Ollama allowing you to run AI models locally :)

How easy is it to create AI agents? without code (or at least minimal!)?

Take a listen as we refresh our memories with n8n.io. Build automation workflows but specifically AI agent-based ones! all this locally or on your infrastructure too!

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Well, I have been trying to settle on some day-to-day tools i.e. OneNote for my note taking (don’t shoot me!) and I did have Todoist for tracking my endless lists of things to either “really do” or attempt to capture for later (yeah right!).

Although there are lots of choice, and that is a good thing, I thought with Todoist I had found the perfect “ToDo” app (or just a large bucket of broken dreams!). No problem with Todoist at all, but I came across Tick Tick and dismissed it a few times.

About 6 months ago I figured I would give Tick Tick a try. Anyway, have a listen if you want to find out how that went.

No affiliates or selling here, just info from my perspective :)

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Unicorn University:

https://applied-ai-services.kit.com/020c6040f2

So, are you “vibe coding”? or just asking AI to help with some code?

Have you come across custom instructions for VS Code? Windsurf and Cursor AI etc. all of their options too.

Provide a preset app requirements and/or product requirements document (PRD) and you can make building apps with AI a bit more predictable and follow your guidelines about the tech stack you want to use, testing approach etc.

Hey, this is a free podcast. However, if you feel you want to support me then check out Patreon. I will have some more detailed deep dives for Patreon members as well as one-to-one sessions.

Or just buy a unicorn a coffee here!

Oh, and yes, I have ended up on YouTube (doesn’t everyone eventually?):

https://www.youtube.com/@justfifteenmins but don’t worry, no ugly face to worry about (yet!).

Unicorn University:

https://applied-ai-services.kit.com/020c6040f2