"use strict";
const {Fragment: _Fragment, jsx: _jsx, jsxs: _jsxs} = arguments[0];
function _createMdxContent(props) {
  const _components = {
    a: "a",
    h2: "h2",
    li: "li",
    span: "span",
    ul: "ul",
    ...props.components
  };
  return _jsxs(_Fragment, {
    children: [_jsxs(_components.h2, {
      id: "retrieval-augmented-generation-rag",
      children: [_jsx(_components.a, {
        "aria-hidden": "true",
        tabIndex: "-1",
        href: "#retrieval-augmented-generation-rag",
        children: _jsx(_components.span, {
          className: "icon icon-link"
        })
      }), "Retrieval Augmented Generation (RAG)"]
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsxs(_components.li, {
        children: [_jsx(_components.a, {
          href: "https://blogs.nvidia.com/blog/what-is-retrieval-augmented-generation/",
          children: "RAG"
        }), " is how you can give LLMs access to your corporate data so they can answer questions with your proprietary internal knowledge."]
      }), "\n", _jsx(_components.li, {
        children: "Ask Steve can be pointed to an API that uses your RAG system instead of directly to an LLM so it can answer employee questions, wherever they are, in the flow of work."
      }), "\n", _jsxs(_components.li, {
        children: ["This example uses the open source vector database ", _jsx(_components.a, {
          href: "https://qdrant.tech/",
          children: "qdrant"
        }), " for storing information and the visual workflow builder ", _jsx(_components.a, {
          href: "https://n8n.io",
          children: "n8n"
        }), " for creating the code that connects OpenAI's gpt4o-mini to the vector database."]
      }), "\n", _jsx(_components.li, {
        children: "An Ask Steve Skill was created that calls the n8n API, which tells the OpenAI LLM to search the qdrant vector database and use the results in its answer."
      }), "\n"]
    }), "\n", _jsx("br", {}), "\n", _jsx("div", {
      style: {
        position: 'relative',
        paddingBottom: '56.25%',
        height: 0,
        overflow: 'hidden',
        maxWidth: '100%',
        height: 'auto'
      },
      children: _jsx("iframe", {
        src: "https://www.youtube.com/embed/taNoHYBT3uk?si=c4aEojkTDNFoiC6O",
        title: "YouTube video player",
        frameBorder: "0",
        allow: "accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen",
        referrerPolicy: "strict-origin-when-cross-origin",
        allowFullScreen: true,
        style: {
          position: 'absolute',
          top: 0,
          left: 0,
          width: '100%',
          height: '100%'
        }
      })
    })]
  });
}
function MDXContent(props = {}) {
  const {wrapper: MDXLayout} = props.components || ({});
  return MDXLayout ? _jsx(MDXLayout, {
    ...props,
    children: _jsx(_createMdxContent, {
      ...props
    })
  }) : _createMdxContent(props);
}
return {
  default: MDXContent
};


Demos


I've had several conversations with very active ChatGPT users recently who mentioned security as a concern, and didn't know that you can run AI/LLM models **locally**. When you run a model locally, the LLM runs on your own computer. The primary benefit of this is that your requests never leave your machine, so all your data is kept private. The secondary benefit is that it's FREE. 

Using local AI models does have drawbacks. First, you need to be able to download and store a large model file (e.g. 3+ GB). Second, you need a reasonably powerful computer to run these models effectively, otherwise they'll be too slow to be of use. Third, the quality of locally-run models typically lags behind cloud-based models - although they are more than good enough for many cases. Finally, despite the local processing, performance can still be slower than cloud alternatives because cloud systems are running in giant server farms. 

There are several applications that make using local models easier. Here are the ones I often point people to because they have simple UIs. Some even enable chatting with your own files.

* [LM Studio](https://lmstudio.ai/)
* [Jan](https://jan.ai/)
* [gpt4all](https://www.nomic.ai/gpt4all)

In general if you download one of these, you'll be given an option to pick a model to download from their library. Picking one of the more recent Meta Llama models or Mistral models that is in the 4GB range is usually a safe bet. You could also try DeepSeek R1 to see what all the buzz is about - using one of these tools is a really easy and secure way to do so without having to worry about where your data is going.

(Picture above is me chatting with DeepSeek R1 in LM Studio running on my MacBook)

-- rajat



Connect to a Local LLM/Model Running On Your Own Computer for Ultimate Privacy and Security


## Use Ask Steve as an AI Writing Assistant in Google Docs and Word 365
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto' }}>
  <iframe 
    src="https://www.youtube.com/embed/l3d9JgCQluw?si=d2c42GNJiYzxSPgD" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>
<br />
I've been using Google Docs more recently to write things for print publications, and have found Ask Steve to be super helpful as a writing assistant and editor. 

It can provide suggestions and commentary, generate lots of alternatives, and more - basically be another set of eyes on what you're writing. 

And since it's right there where you're writing, it's 100x more useful than bouncing back and forth to ChatGPT or Gemini. 

You can either ask for whatever you want, or use one of the many writing-related Skills from the library. 

It also works in Word 365.

Give it a try! - rajat


Use Ask Steve as Your AI Writing Assistant in Google Docs and Word 365


### Read the full article [here](/hacker-news)!


How I Built a 1-Click AI Comment Generator for Hacker News


## MindStudio and Ask Steve - Better Together!
Watch this video to see how to use the new [MindStudio](https://mindstudio.ai) integration in Ask Steve.
*This requires the Ask Steve Premium Plan in order to call a [Webhook](/docs/webhooks)*
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto' }}>
  <iframe 
    src="https://www.youtube.com/embed/5p8kADzf9YI?si=MS0OZlFOYLDazUdV" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>

MindStudio and Ask Steve - Connect Your 1-Click AI Assistant with Backend AI Workers and RPA


## Get help with formulas in Google Sheets
Need help with formulas? See how you can add your own Help button to Google Sheets that you can ask for formula help!
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto' }}>
  <iframe 
    src="https://www.youtube.com/embed/dGzMJKYJF2o?si=GR2gLbOsEPzTNLWe" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>

Use Ask Steve's AI Assistant to get Help in Google Sheets


## Ask Steve for Sales & SDRs
You have numbers to hit and Ask Steve can help! With these Skills and Buttons, you can save time extracting leads from web pages, researching companies and prospects, qualifying leads, crafting cold outreach messages, and writing ongoing comms with prospects.

We added 2 Skills Buttons to LinkedIn profile pages so you can see how you can qualify leads and write customized outreach letters with 1 click. Customize them and make them yours!

**Ask Steve is your secret weapon to hit your number and get your bonus!**
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto' }}>
  <iframe 
    src="https://www.youtube.com/embed/XNfZKJnN_ec?si=zJ-YQ7iuckQFUBxr" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>

How Sales People and SDRs can use Ask Steve's 1-Click AI Assistant


## Retrieval Augmented Generation (RAG)
* [RAG](https://blogs.nvidia.com/blog/what-is-retrieval-augmented-generation/) is how you can give LLMs access to your corporate data so they can answer questions with your proprietary internal knowledge. 
* Ask Steve can be pointed to an API that uses your RAG system instead of directly to an LLM so it can answer employee questions, wherever they are, in the flow of work. 
* This example uses the open source vector database [qdrant](https://qdrant.tech/) for storing information and the visual workflow builder [n8n](https://n8n.io) for creating the code that connects OpenAI's gpt4o-mini to the vector database.
* An Ask Steve Skill was created that calls the n8n API, which tells the OpenAI LLM to search the qdrant vector database and use the results in its answer.
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto'}}>
  <iframe 
    src="https://www.youtube.com/embed/taNoHYBT3uk?si=c4aEojkTDNFoiC6O" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>

Connect Ask Steve to your AI & LLM Powered Retrieval Augmented Generation (RAG) system

Give your employees access to your corporate knowledge-base in the flow of work

Retrieval Augmented Generation (RAG)

Read more about
⚡Demos

Connect to a Local LLM/Model Running On Your Own Computer for Ultimate Privacy and Security

Use Ask Steve as Your AI Writing Assistant in Google Docs and Word 365

How I Built a 1-Click AI Comment Generator for Hacker News

MindStudio and Ask Steve - Connect Your 1-Click AI Assistant with Backend AI Workers and RPA

Use Ask Steve's AI Assistant to get Help in Google Sheets

How Sales People and SDRs can use Ask Steve's 1-Click AI Assistant

Connect Ask Steve to your AI & LLM Powered Retrieval Augmented Generation (RAG) system

Give your employees access to your corporate knowledge-base in the flow of work

Retrieval Augmented Generation (RAG)

Read more about ⚡Demos

Connect to a Local LLM/Model Running On Your Own Computer for Ultimate Privacy and Security

Use Ask Steve as Your AI Writing Assistant in Google Docs and Word 365

How I Built a 1-Click AI Comment Generator for Hacker News

MindStudio and Ask Steve - Connect Your 1-Click AI Assistant with Backend AI Workers and RPA

Use Ask Steve's AI Assistant to get Help in Google Sheets

How Sales People and SDRs can use Ask Steve's 1-Click AI Assistant

Read more about
⚡Demos