238 | Win More Business with Less Effort - How to Write Winning Proposals in 15 Minutes Using AI

4 Nov 2025 · 40 min

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

Leveraging AI Podcast - Episode 238 Summary

Episode Title Win More Business with Less Effort - How to Write Winning Proposals in 15 Minutes Using AI

Episode Description This episode features a solo masterclass by Isar Meitis, focusing on how business leaders can leverage AI tools (such as ChatGPT, Claude, and NotebookLM) to create effective proposals in a fraction of the traditional time. The episode covers strategies for transforming meetings into polished proposals, handling RFPs, and writing emotionally intelligent cover letters.

---

Key Concepts

The Importance of Proposals

  • Proposals are essential for acquiring clients; without clients, businesses cannot thrive.
  • A successful proposal must:
  • Clearly articulate client needs.
  • Connect emotionally with the client, as decisions are often emotionally driven.

Understanding Client Needs

  • Proposals can be based on:
  • Direct Communication: Meetings, emails, chats.
  • Requests for Proposals (RFPs): Detailed documentation from clients outlining requirements.

Proposal Writing Process

  1. Capture Client Communication:
  2. Use tools to record meetings and calls (e.g., Zoom, Fathom).
  3. Transcribe verbal communications for easy reference.
  1. Using AI for Proposal Creation:
  2. ChatGPT: Create a custom GPT (Generative Pre-trained Transformer) to draft proposals based on transcripts.
  3. Claude: Best for generating complex financial proposals in Excel format.
  1. Integrating AI into Workflow:
  2. Develop custom GPTs or Claude projects that streamline the proposal writing process.
  3. Include templates and instructions for the AI to follow for consistency and personalization.
  1. Review and Edit Outputs:
  2. Use the generated proposals as drafts, making necessary adjustments to align with client specifics.
  3. Maintain a side-by-side editing format (e.g., Canvas in ChatGPT) for easy modifications.

Enhancements for Proposals

  • Emotionally Intelligent Cover Letters:
  • Attach personalized cover letters to proposals, capturing emotional connections with the client.
  • Use AI to draft these letters based on the proposal content and client research.
  • Visual Tools for Clarity:
  • Tools like NotebookLM can create mind maps or audio summaries to simplify complex proposals, enhancing client understanding.

Dealing with RFPs

  • Understand the structure and requirements of RFPs thoroughly.
  • Establish clear guidelines for the AI to follow, ensuring it bases its responses strictly on provided documents.

---

Key Takeaways

  • Efficiency: The integration of AI reduces proposal writing time from hours to approximately 15 minutes while maintaining quality.
  • Customization: Tailoring proposals to client needs with emotional intelligence significantly boosts chances of winning business.
  • Automation: Leveraging AI tools for routine tasks can free up time for more strategic activities within business operations.

---

Additional Resources

  • [Master the Art of End-to-End AI Automation](https://multiplai.ai/advance-course/)
  • [AI Business Transformation Course](https://multiplai.ai/ai-course/)
  • [YouTube Full Episodes](https://www.youtube.com/@Multiplai_AI/)
  • [Connect with Isar Meitis](https://www.linkedin.com/in/isarmeitis/)
  • [Join Live Sessions and Events](https://services.multiplai.ai/events)

---

Conclusion In this episode, Isar Meitis emphasizes the transformative power of AI in proposal writing, illustrating how business professionals can leverage these technologies to improve efficiency and win more business. From initial client contact to the delivery of polished proposals, AI can dramatically streamline processes, making it a valuable asset in today’s competitive landscape.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Hello and welcome to the Leveraging AI podcast, the podcast that shares practical ethical ways to leverage to improve efficiency, grow your business and advance your career. This is Isar Maitis, your host. And in this episode, we're going to dive how to write winning proposals in order to get more business while investing significantly less time. If you have been listening to this podcast for a while, you know that I've been saying multiple times that a business can run without HR, it can run without marketing, it can run without accounting, it can run without a lot of things that exist in most businesses, it cannot run without clients.

0:32Because if you don't have clients, you don't have a business. You could be a good nonprofit, but you cannot run a business without clients. But to get clients, you need to be able to write proposals, not just proposals. You need to write proposals that actually win you business. So what does a good proposal look like? So the first thing you need is to understand what the client exactly wants. You also want to connect with the client emotionally, because if you can connect with them emotionally, there's a significantly higher chance they're going to buy from you because humans as living creatures make most of the decisions based with emotion, and then we justify it based on parameters.

1:05And so if you can, A, present your case in an easy way to follow, connected with the client's needs and connect with them emotionally, even better. Now, as far as what I said, making it easy to follow and understand is a big deal of writing a successful proposal. And there's actually two types of proposals, or there's probably more, but two big categories of how you write a proposal or how you understand the client's requirements. One is based on communication with a prospect or an existing client for an additional work. And that could be based on meetings and demos and emails and chats and different ways of communication.

1:38So you can understand exactly what they need. The other option is an RFP. In bigger projects, the client will write a request for proposal, which will clarify exactly what they need, how they want the proposal written, and so on. And then you need to write them a proposal based on that. And we're going to cover both of these options in this episode. Now, either way, writing good proposals is really time consuming. And in many cases, it's not straightforward in connecting the dots of exactly what the client wants, or whether through an RFP or through the different means of communication, and writing a proposal that will capture all of that in an effective way.

2:13Now, also in many proposals, you have two aspects. You have the written proposal, which would be a Word or a PDF or a Google Doc. And there's the financial component that in most cases is going to be Excel or a Google Sheets, and we're going to cover both these as well. We are going to start with the first option of capturing the client's requirements, which is communication with the client and not an RFP, because I believe that most of you write most of your proposals based on that. So what kind of communications can you have? This could be meetings in person, either in a showroom or at your offices, but this could also be a Zoom call or emails or Slack or whatever means you use in order to chat with your clients or a combination of all of the above.

2:52So what you need is you need to capture all of these communications. So the ones that are written are very easy. You have their either in some kind of a chat platform or on your email, but the communications that are verbal, you have to record. You can either record them if you're in a Zoom call or Teams calls or Google or whatever other means you have your virtual communications. Or if it's in person, you can still record it in person. You can either do this with several different types of applications on your phone or buy a hardware product such as Plod that is the size of a credit card, gets stuck to the back of your phone, and you can then record calls in person as well.

3:26But either way, you want to capture these calls. I do all of my communication with my potential clients and prospects on Zoom. And so I record the Zoom calls and capture all the information. I use a tool called Fathom, which is a tool that can record, transcribes, and get notes from every single meeting. And there are many other tools that do the same thing. I really like Fathom for many different reasons. I find their summaries to be very effective and it's very easy to use and it saves all of the recordings in the history so I can go back and take a look at exactly what happened so I can get the context of what was said and not just the summary.

3:58But all of that doesn't really matter for the proposal because when I come to write the proposal, I don't really care. When I'm going to fathom, I have the summary of everything that happened as far as bullet points as a summary, but I also have a transcript. All I have to do is click on copy transcript and I'm good to go. what do I do with the transcript that I copied? Well, I take the copy that I have and I go to ChatGPT. And inside of ChatGPT, I have a custom GPT that I have created, which I'm going to show you exactly how to create yourself. And in that custom GPT that is called Multiply AI Training and Education Proposal, all I have to do is paste the transcript.

4:34That's it. So then I click go, and it is going to read it, and it's going to write a proposal. Now, before we jump into the proposal, I want to say two words about custom GPTs. Those of you who don't know what custom GPTs are, there are these mini automations that you can build inside of ChatGPT. By the way, as I'm speaking, it is writing the proposal. So those of you who are watching this on YouTube can see the proposal being written, but I'm going to read snippets out of it to you once I'm done. But these custom GPTs are these mini automations that you can create. You can get to them on the left side menu after you have the new chat and search and library and so on.

5:05There's a section called GPTs. If you haven't created any, you're just going to have a button called Explore. I have about 30 plus of them that me and other people in my company use regularly to do almost everything in our business, run through different custom GPTs. So this particular GPT writes proposals based on a transcript. So let's see what we have right now. We have a proposal written, and you can see that it shows up in this little interesting square thingy versus the regular chat. And you can see that it has an edit button inside the top of that square. This is because I requested the custom GPT to write this in what ChatGPT calls Canvas.

5:41So if I click on edit, now what I have is I have a side-by-side setup where I have the chat on my left and I have the proposal on my right. And on the top of the proposal said AI training and consulting proposal for We Are Sassy, which is a company I completely made up, but I really do this with every single proposal that I write. And then it says provided to John Smith, founder of We Are Sassy, provided by Isar Mehdi, CEO at Multiply, date November 3rd, 2025. Introduction. This proposal outlined a tailored AI training and implementation program for We Are Sassy. The initiative is designed to accelerate your journey into the AI-enabled business operations by identifying and unlocking high ROI use cases, streamlining proposal writing and lead generation, and building internal capacity through custom live training sessions.

6:26And then it goes to objectives. Identify and implement high ROI, AI use cases within the next 90 days, build internal capacity and automate core workflows, such as proposal writing and lead generation, establish baseline AI capabilities, et cetera, et cetera. It goes through all the things. Then he talks about why AI training is urgent, and he touches on exactly the topics that they mentioned. As you can imagine, in this particular case, it was lead nurturing and proposal writing. Why work with multiply? So there's several different bullet points. Why to hire me, basically, to be the one delivering the training?

6:58And he touches on a very important point that, again, are tied back to their needs. And then he talks about training formats and recommendations. So it gives several different options of formats, and he talks about why they should do each one. So there are great justifications. I'll read a short snippet out of that. Online workshops, in parentheses, use case specific. We recommend a series of focus two-hour workshops delivered live via Zoom. These workshops are designed to address specific bottlenecks and high-value opportunities, in parentheses, proposal, automation, lead generation, et cetera.

7:27Bullet point number two, include live demos and hands-on experiences tailored to your business. Bullet number three, involve core team members and freelancers, regardless of location. This particular real client, again, I've changed the name, has a relatively small in-house team and a very large team of freelancers. And the idea was to create online workshops so other freelancers from all around the world can participate in these workshops. And then it goes to other options that they could do instead or in addition to the two-hour workshops that were suggested. Then it goes to a roll-up plan. What happens in November?

7:59What happens later on? What happened in December? What happens next year? Then it goes to pricing. Then it goes to data preparation and confidentiality, like what kind of data they need to provide to me, and so on and so forth. It's a really long and highly detailed proposal. Now, I know that those of you who've never done anything like this go, holy crap, I want something like this. How the hell, did you just do this? And so two things before we dive into how I did this. The reason I do this within Canvas, which again is this format where I can see the document on the right and the chat is on the left, is because now I can edit the document.

8:32If I want to add another objective or I want to change the objectives, I can click inside of it and I'll just type. And I can now type whatever I want inside this document. I can delete components. I can add components. I can reorder a component just by copying and pasting. Think about it as having Microsoft Word or Google Docs within ChatGPT. Now, the other thing that is very helpful, because it is within ChatGPT, I don't have to do the editing manually. So one of the things here, it says, why work with multiply? And it has already existing four bullet points. The third one is combines business acumen with AI expertise to focus on real practical ROI.

9:06Sounds great, but let's say I want to say more about it that is related specifically to their business. I can highlight this section. And then when you highlight anything within Canvas, it pops up a little pop-up thing. And one of the options there, it says, Ask Chat GPT. So if I click on that, now it is going to work related only to the highlighted area. In this case, a specific bullet point. And I can say, please provide more information about this aligned with the needs of the client. And then it is going to rewrite just this one bullet point. As you can see now, or those of you who can't see, I will read it.

9:45There used to be just one sentence, and now it's an entire paragraph with more information related specifically to the conversation that I had with the client. So this is the huge benefit of Canvas. And basically, because I've defined exactly the format that I want, so it has numberings and headings and bullet points, exactly the way I usually write proposals. Once I'm done editing it here, I'm basically done. All I have to do, there's a copy button on top. I copy it, paste it into my regular template, either in Word or Google Docs with my header and footer and so on, payment terms, all the stuff that is standard, and I'm done and I can send a proposal.

10:20But how did it happen? How does this thing work? So I'm going to close the canvas for a minute and I'm going to go back to the custom GPT that again, you can see here on the left and I will show you exactly how it works. So if I go to edit GPT, how does a GPT work? By the way, I've done an entire episode about how to create custom GPTs, and that's not going to be the focus of this one. If you want to check the custom GPT episodes in which we dive exactly how to develop them, you can go back to episode 175 that is labeled, Stop Wasting Time, Automate Repetitive Task with Custom GPTs. So you can go back and check that one again, episode 175.

10:57But now let's dive into this custom GPT. Its name, as I mentioned before, is Multiply AI Training and Education Proposal. The description says, writes proposals based on a template and meeting transcription. And then there's the actual instructions. By the way, the name and the description doesn't matter. It's just for you to know what it does. So if we look at the instructions, the instructions are where the magic actually happens. And it says the following, you are an expert proposal writer. Your goal is to write clear, easy to follow and attractive AI training proposals. Cool. Now you obviously write your own thing, however you want it.

11:26Now I define the inputs. So the inputs include, and I'm quoting, use the master template of AI training and consultation proposal in your knowledge base as a template. And then it gives different options of the things that I do. But then it says the following. Each proposal that you write will include one or more of these training options. You need only to include the components that were discussed with the prospect in parentheses based on the transcripts and emails I will provide to you. Also, you have access to the Multiply AI Services brochure. Every time you will write a section from the proposal about a specific type of service, you must add a short description of the relevant service and its benefits from this brochure.

12:07Combine it seamlessly with what was discussed in the call. And in parentheses, it says what the prospect or client said and how they said it. Then it talks about what is the second input. So the first input was documents that I'm providing it. I'm going to touch on that in a minute. The second input is client meeting transcript and or email communication with the prospect slash client. The user will provide you with a full transcript of a discovery call with a potential client, the transcript, who the client is and what do they do, what they are trying to achieve with AI, team structure and distribution, any strategic initiatives such as upcoming sprints, rollouts, goals, etc.

12:41And there's a bunch of those you obviously need to write your own. Yours is going to be very, very different than mine. I provide AI training consultations to companies and workshops to organizations. And so this is my instructions, but yours could be very, very different. And then I define the output. The output is using the template proposal as a base, generate a customized client-ready proposal. The final output should, and now there's a list of things, update the title and headers, because I have a generic title in headers, right? So include client's names instead of, and the template has company name in bold letters in order to tell it what to replace.

13:16Update the recipient names and relevant dates, adjust introduction and objective sections, et cetera, et cetera, et cetera. And then it tells it all the things it needs to update, all the use cases that it needs to bring up. And then it also suggests to customize the data prep section, include any use cases and or documents they plan to share, Jira stories, test cases, anything that they're going to provide to me that was discussed in the call or in the emails. It also suggests to add client-specific benefits and ROI. And the two things that I'm suggesting is reflect their business objectives, e.g.

13:48speed of sprint execution, team enablement, et cetera, whatever they're going to mention. And number two, emphasize alignment with their transformation goals. And you've seen that in the proposal that it connects back to what they said in the call and in the emails. I have really exciting news for you today. The biggest thing that everybody's after with AI is automating different tasks in your business, right? This is the holy grail that everybody's chasing. This is why these companies are worth many billions of dollars because they promise to automate stuff that otherwise will require people to do the work.

14:20Well, most people hear about this and drool about this, but they don't have a clue how to actually build these automations. Many people who have taken my course, the AI business transformation, thousands of people at this point learned the basics of how to prom properly and how to use different AI tools, but to learn how to actually effectively build automations that can actually automate tasks across any aspect of the business is something that requires the basic training that I've delivered and then going beyond. And this is exactly what we are starting to deliver. Starting on the first week of December, we are launching a course that will teach you everything you need to know to automate stuff in your business, how to connect your existing tech stack, your CRM, your ERP, email, marketing platform, whatever aspect you want together with advanced AI capabilities to create workflows that automate more and more aspects of your business.

15:12Think what that might be worth. This can dramatically accelerate more or less every single thing that you do while dramatically reducing the cost, freeing people to do more important jobs than the mundane stuff they're doing every single day. So everybody wins. You have happier partners and employees. you run faster and you drive more business, all while investing less money than you're investing right now, which means regardless of the size of your business, this is an incredibly valuable offer. And we're doing this for a price that makes it a joke. So if you want to learn more, we are launching this course again, the first two weeks of December.

15:50It is going to be two weeks on Mondays at noon Eastern. So you can join us at noon, December 1st, and then December 8th. And And in four hours, we will take you from I know how to prompt and I know how to use AI at a basic level to having all the tools, including templates that I have personally created and that I use for myself and for my clients that you are going to get for free that will get you and save you hours of figuring it out on your own. So come and join us in December. By the way, if you don't have the basic knowledge and you're saying, this sounds a little above my head, I need to get the basics first.

16:20We are launching another cohort of the highly successful AI business transformation course that I've been teaching for two and a half years since April of 2023. And we're updating it every single time we're running it. And we've run it at least once a month since. So it's been updated many, many times. We are launching another cohort of that at the end of January. So if you want to start 2026 with the right foot forward, knowing either the basics, if you're a beginner, or the more advanced automation capabilities, don't miss this opportunity to start 2026 with the right foot forward. I'm not sure what we're going to teach the next cohort of this.

16:56So this is an amazing opportunity to finish this year strong and start the next year even stronger. All of our courses sell out, literally 100 % of them. So don't wait. There's going to be a link to both courses in the show notes of this podcast. just click the link, check the landing page. It has the syllabus and everything you need to know about the course. And you can then start automating your business and moving significantly faster next year. And now back to the episode. Then there's a few additional things that are very specific as far as the formatting and how I want it, what to include and not to include.

17:28And in the very, very last sentence, it says, write the proposal in Canvas, which means it is going to write it in that format that allows me to edit the proposal and work in collaboration with ChatGPT on making it perfect. So what else is in the CustomGPT? There is the knowledge base. What is the knowledge base? The knowledge base is where you can give the CustomGPT additional information. In this particular case, there are two files. The first file is the master template of AI training and consulting. And that template has literally every single service that I provide, which I will never actually use in a real proposal, but it allows the AI to then pick and choose the different services that was defined in the conversation with the client.

18:06So as you heard in the instructions, I tell it to pick and choose just the component that it needs. The other document is the Multiply AI services brochure. This has a much longer description of the different services that I provide, and the AI knows how to pick from that in order to describe the services and the benefit that they provide to the client based on what they said and based on their needs. And then the final section here, it says capabilities. And in the capabilities, you can choose different options like web search, canvas, image generation, code interpreter and data analysis. And the only thing I checked is canvas because it's the only thing I need.

18:37And that is it. If you take this exact concept and make it your own, meaning combine your proposal template and your information about the services and product that you sell with instructions on how to use a transcript or any other means of communication with the client, you'll be able to write your own proposals. something that used to take me about two hours, takes me about 15 minutes right now with most of the heavy lifting actually done by AI. But as we mentioned, this is just one aspect of the proposal. You also need the financial side of it. You want to deliver an Excel, sometimes a very complex Excel with multiple tabs and multiple formulas that are connecting it together.

19:15And the best way to do that is actually in Claude. So Chachipiti is great and it's even okay at running formulas in Excel. But once you start getting into really complex Excel files, especially with multiple tabs, Cloud becomes a much, much better tool, at least as of right now. So let's look at a similar example in Cloud. So what I've done here in this particular one, and I'm not going to run it in real time because it takes it a while to create. So what you can see here is a audio visual proposal generation. And the reason I chose that is A, because I have a lot of client in the smart home industry, but also because it has a gazillion different options and different things.

19:53And what you can see here is that I'm looking at a page out of a multiple page Excel file that was completely created by Claude. And all I had to do is a very similar thing. You can see that I uploaded a document that's called Client Showroom Meeting Text, which is basically a text transcript that was taken from a hardware device inside the demo room. And it captured the entire conversation with the client. And all I did is just that and click go. and it said, and now Claude said, I've reviewed the client showroom meeting transcript before I begin creating the Excel proposal, I need to clarify several important items to ensure accuracy.

20:25And it's going to ask me multiple questions. Why does he ask me questions? Because the instructions that it has inside the Claude project, ask it to ask it clarifying questions. So it says, okay, what are the names of the clients? What is the theater configuration? Are we looking at the theater system tier or other things? So different options of how that may look like. What kind of audio details it is going to include? It provides me several different options based on the conversation. What is the video wall size? Because two options were mentioned in the conversation. What is the speaker selection?

20:54So there's several different options for that. Room by room count. So what is going to be the number of speakers in each and every one of the rooms? Because that may have not been finalized or agreed upon in the meeting transcript that it has. TV location and sizes. Vinyl player location. So a lot of very relevant questions that it needs to answer in order to do the proposal properly. And it goes on and on and on. There are 14 different questions. Because I wanted to do this as a demo and I wanted to run this quick, I basically said act as an AV expert, answer all these questions on your own and create the Excel proposal.

21:25I will obviously never do that with a real proposal. I will actually answer all these questions in detail, most cases by voice. So I click on voice type. I voice type most of what I do right now and just answer all of those questions. And then Claude went ahead and started working on the proposal, wrote all the different components and provided me with this highly detailed Excel with the dining room and kitchen and pantry and what are the speakers and what are the control panels and everything that exists in every single room. And this is, again, just one page out of like six or seven pages, each and every one of them with different sets of systems that all feed back into this Excel file.

21:57And you can see up here on the top, because I connected it with my Google Drive, I have a drive button. And if I click on that, I can download it straight into Google Sheets. You can do the same thing by connecting Claude to SharePoint and then download it as an Excel file straight into SharePoint. Is this perfect? No. Does that get me 85 % of the way there? A hundred percent. And sometimes it gets me 98 % there with all the different tabs, all arranged, all in the format that my client is using to provide answers to their clients. And it all is generated instead of in a few hours, in a few minutes.

22:33And again, all these minutes are spent by Claude and not by you. So all you have to do is upload the transcripts or any other notes or communications that you had with a client, and it will write this proposal. And then the following step is very similar, right? It is combined into a proposal, as you have seen before. So I can actually take the output from this, upload it to the other custom GPT, or create a similar concept inside of Cloud. So in Cloud, custom GPTs are called projects, and this works exactly the same. So now let's take a look at how this is created. So this thing, if I go to projects in Cloud, and it looks like a little folder thing on the left menu in Cloud, you can see that I have AV proposal generation, and you can see that it has multiple files connected to it, like showroom meeting agenda, template, round examples, blah, blah, blah, blah, blah, different PDF files of different things and how the output should look like.

Read the full transcript

23:22And then it has instructions just like we have seen before. So Cloud project AV proposal generator, converting client consultation transcripts into detailed Excel proposal. And then it has project overview and knowledge requirements, and core workflow and the different phases. So that's phase one, transcript and analysis. And it has several different phases. You can see that in phase one, it goes through transcript analysis and clarification. That's where it's going to review the transcript and ask me questions. And there's a bunch of them. And then there's phase two, the actual Excel generation.

23:50But inside the Excel generation, there are multiple steps. And it's going to generate that highly detailed Excel file. So quick summary before we continue. Once you create either a cloud project or a custom GPT in ChatGPT, you can then take the transcript or communications that you had with clients and prospects, upload them and get a proposal that is almost completely ready, both the pricing side of it in a highly complex multi-tab Excel file and the written proposal that will go with that. But now let's talk for a minute about working with RFPs because some proposals have that. By the way, I have two really interesting bonus things in the end.

24:29So even if RFPs are not interesting to you, stick around. I promise you it will be worthwhile. But let's talk about RFPs for a minute. So in this particular case, I went to a government website and downloaded a construction RFP. It's called Solicitation Offer and Award. It has a solicitation number. It has multiple different pages, multiple different documents, and exactly what it's going to be. and it is for renovating the Indian Education Acquisition Center in Indian School Roads with 352A in Albuquerque. That is the actual, what the solicitation is about. And I'm not going to read all of it to you, but it's a really, really long, ugly government generated PDF that is an RFP.

25:04Most RFPs are like that. It has multiple, multiple details across multiple pages, and in most cases across multiple documents as well. So how do you use that to write the document? The first thing to write the proposal, the first thing you need is to understand exactly how the document structure works or the documents work and to get answers on whether you can even bid on that. Do you have the right expertise? Do you have the right licensing? Do you have the right insurance? Do you have et cetera, et cetera? And that goes into the process of how do you work with a document inside of these AI tools?

25:38And this is what I'm going to teach you right now. So the trick here is to have a detailed understanding of what's in the document while reducing the hallucinations that AI has when it's reading documents and reducing the amount of time it will take you to verify that it's not actually hallucinating. So you can see that I started a conversation with ChatGPT, by the way, on this particular process of working with documents as of right now. So November of 2025, ChatGPT works better than Claude and Gemini in the capacity that I'm going to show you. So I've uploaded the multiple documents. It's just one PDF, but there's multiple documents in it.

26:11And then, or you can upload it, by the way, as separate documents. And then I wrote the following. You are an expert government proposal writer who specializes in writing winning proposals for government construction RFPs. You will be provided with a document as an attachment and several questions you should answer. But then before I actually asked the question, I added the following component, which is the most critical aspect. All of this is what I really want you to remember from the whole RFP thing. So a few important rules for this process. One, you must use information only from the document I attached.

26:42Two, do not use any other sources for any step of this task. Sounds redundant, but it's not really because it covers the two sides of that story and it increases the chances that the AI will use just the information that you provided it. Number three, and this is where it becomes very, very critical. If you cannot find information that is directly related to the questions you are asked in the original document, simply respond with, and then in quotation marks, I put insufficient information. You can put whatever thing in that quotations. You can say information not available or NA or whatever you want.

27:15Why is this so important? Because these AI systems are built to please. They want to provide you an answer. You're going to ask a question, you're going to get an answer. Even if the information doesn't exist in the document, in many cases, it will make it up in order to answer your question. However, if you provided instructions, what it's doing, it's actually better than just providing a random question because it is following your instructions, which is what it's built to do. So if you're telling it what to say if the information doesn't exist, there's a much, much higher likelihood it is not going to make it up.

27:42And it's actually going to say that the information doesn't exist in the same exact wording that you requested. And then number four is for every question you are asked about this document, documents, you must include in every segment of your answer a citation in the following format. And then I have squiggly brackets, document name, comma, page number, comma, heading, comma, exact quote. Sometimes I ask for it in a table instead of just this format. But either way, you will see if I scroll down and I look at the answer, it answers. So I ask it the question about what is the performance period of this job once awarded?

28:17And you can see that it tells me performance period 365 days in parentheses, 365, calendar days from notice to proceed. and then it shows me in squiggly brackets exactly what I said. It gives you the document number, the page number, the name of the section, and then it actually gave me the exact quote. It also gave me another one because it found another reference to this in a separate document in that file. So you can see there's another place on page 16 in section F.1, commencement, prosecution, and completion of work. So you can see that it is very, very detailed in providing me these citations.

28:52Why are these citations important? They're important for two reasons. Reason number one, it forces the AI to ground itself in something that actually exists in the document. Reason number two, it allows me to very quickly verify that what the AI is telling me is actually correct. Because all I need to do right now is to go to the original document and copy and paste the quote that it has given me and hit command F on the document and see if this quote actually exists. So let's try this. I'm going to hit command F. I'm going to paste it. And you can see that it actually exists exactly where it says.

29:23So now it highlights that section. I can see that it's really in the page that it said. I can see it's in the section that it says. And now I know this information is accurate. And I will do this for every single information that the AI gives me. And yes, it will take me another three minutes to do or five or 15 or 25, doesn't matter. But I will know that all the information that I got, that finding it manually across several different documents will take me hours. Now I have the answers to all my questions related to this RFP. And then I also can verify that information is real. And once I have all of that, then I will take the outputs of this process of the questions that I asked, and then I will upload them to a very similar process to what I showed before.

30:00In this particular case, the instructions of the custom GPT, instead of relating to information that is coming from a conversation, it is going to be based on the information from the RFP. But other than that, it is a very, very similar process. So a quick recap on what you need to do based on what we've done so far. And then I'm going to give you the two bonuses that I promised you. One is you must record all your calls with the client and or prospect, whether it's in person or on Zoom or over your phone. You need to record these calls for all your salespeople in order to do this process effectively.

30:29And if you're not doing it right now, then start doing it. Two is you need to create a solid, comprehensive template on both the Excel side and the Word slash Google Docs side, because that is going to be the baseline for what the AI is going to use. Then you're going to start in a regular chat. So not in a custom GPT and not in a cloud project, you're going to start in a regular chat and explain to it what you're trying to do, including how you read the template and how to capture the customer's voice and tone and their needs and their fears and all of that. And after you work with the AI in the conversation, at the end of that, when it is able to produce a proposal based on the conversation you had with it, You ask the AI to generate the actual instructions for the custom GPT and or cloud project.

31:11You then go to the cloud project or custom GPT. You upload the relevant files. You make sure that they are referenced correctly by the instructions. If not, you need to add that manually. You make final tweaks. You test it. You iterate until you got it absolutely working the way you want, in the format you want, in the length you want, and so on. And you are good to go. Once you have that, all you need to do to get a proposal written is to give either the cloud project or the custom GPT, the call transcript of whatever calls you had, chats, if you had any emails, if you had any and any other means of communication that you had.

31:41If it's an RFP, you need to upload the RFP plus the questions that you asked the AI and the answers that you got from the AI. And make sure that in the instructions, it says to create it in Canvas or in cloud, it's called Artifact, but it's the same thing. It's the side-by-side setup where you can make edits. So make sure that your instructions mention that. And that's it. When you get the output from the AI, you can make changes manually or together with the AI, as I showed you before, by highlighting a section and asking the AI to provide additional details, do additional research, whatever it is that you require to make the proposal better, copy and paste it into your template, and you are good to go.

32:16Now, I promised you two small important bonuses. So bonus number one, in many cases, you want to write an email or a cover letter to go with a proposal. You're not just sending it. You want to add something to it. And over there, it is even more important to capture the emotional aspect and to make a connection on the emotional level with the potential client. And I have a custom GPT to do that as well. So let's take a quick look at one of them. This is an example I've actually done to Truist, which is a local bank here in Florida. They're actually in other places as well, really big bank. And I will show you straight the instructions of what it does.

32:46But what it does in this case, I already have the proposal. So I finished step one and you can see this is a custom GPT. And in this custom GPT, what I have is the instructions that are actually very simple and straightforward. And they say, you're an expert sales communicator specializing in capturing the attention of potential clients with proposal cover letters. Your task is to review the proposal that is uploaded, plus the information about the client, and to use Canvas to draft a perfect cover letter that would personalize to the client and will use the Truist bank spirit and tone. And again, in this particular case, this is an example about Truist, but you will obviously make it about yourself.

33:21But then the instructions are actually very simple. Here are the instructions. One, ask the user to upload the proposal. Once you have the proposal, ask the user for a link to the client's website. Once you receive the link to the client's website, please perform a thorough research on the client. Don't just look at the homepage of the website, but dive deeper to their About Us page, product and services pages, and learn as much as you can about them. Number four, check the information in your knowledge base. You will find the following about Truist and Truist stories. Use these in three ways.

33:544.1, learn about Truist culture and what we stand for. 4.2, learn about the Truist brand and what we believe in. 4.3, learn the style in which we communicate. The entire language comes straight from our website. I would like you to capture the style in which we communicate and use it in the cover letter. And there's several different sections. Number five, it asks it to find parallels and connections, emotional connections, between what the client does and the way the client communicates and the core values that they have to what Truist has on their website. And then it writes the cover letter in Canvas, which I can edit and then use as the cover letter.

34:31So it helps me find emotional connections and alignment between the client and myself without actually having to do this manually every single time. And this makes a very, very big difference when people feel that you actually invested the time in learning about their brand, in learning about what they stand for, and that what you stand for is actually well aligned with it. This definitely helps get the attention. It will allow people to be in the right mindset, in a much more positive mindset when they review the proposal. And then the last thing that I want to show you, in many cases, proposals are really highly complex and complicated and not easy to follow despite all of our attempts.

35:05As an example, those AV proposals that I mentioned before, in many cases has multiple rooms and in some cases, multiple buildings and in each and every one of them, multiple systems. So even if you're very organized and trying to explain everything, it still becomes a very complex proposal. And what I want to show you is two ways on how to make it easy for the clients to understand what's going on. Both of them are happening inside of Notebook LM, which is a free Google tool that you can get to by typing notebooklm.google.com. This tool does a gazillion other things amazing, but in general, it's just very, very good at aggregating and summarizing and presenting information from multiple sources in a way that is easy to consume, which is exactly what we're trying to do with the proposal.

35:45So what you can see here on the left is that I've uploaded, first of all, three different PDFs, which are the different components of the proposal, the AV infrastructure, security system, and lighting control. Again, in your case, that might be very different things. And then there are multiple URLs to different components and systems that are included in this proposal. You can see that it gives me a general summary of what this thing is. But then on the top right, you have different options. You have audio overview, video overview, mind map reports, and so on. And what I've used here, just to show you an example, the first one is I'm using the audio overview.

36:17And if I hit play, you're going to hear two different people explain about this proposal. Sure, for a modern home. Exactly. Think of it like the, well, the blueprint for making a house truly smart. It covers everything. That's right. It's not just, you know, a list of TVs. It details the equipment, sure, but also all the pre-wiring needed for AV, security, even lighting control. It's a whole package. A comprehensive plan for a fully connected residence. Awesome, right? So then it goes, this is a 22-minute podcast that explains all the different components of this proposal that your client can listen to instead of you having to do the meeting, or it could be in any language you want, which is very helpful.

36:52So it's just a great way to present the overall really detailed proposal. But what is actually even cooler in these kinds of really complex multi-level, multi-layered proposals is actually to use a mind map. So if I click on the mind map, you can see, those of you who don't know what a mind map is, it's this multi-level brainstorming ideation tool. But in this case, it's just taking the information straight from the proposal. So you can see on the left, it says project proposal. And the first thing it does, it has three different components, AV infrastructure, security system, and lighting control system.

37:20I did not create this. It understood this on its own. So let's click on the AV infrastructure. And then it opens first floor, second floor, exterior, cable and hardware, labor and costs. So either broken down by the components, broken down by cost, or broken down by the floor of the house and or exterior. So let's go to the second floor as an example. So So now I can see on the second floor, it shows me the rooms, office guest room two, guest suite three, and then WIC. So let's look at the office and guest room. And if I click on that, it shows me all the different components that is going to be there.

37:52And it's like 25 of them. If I click on one of those, so let's say Leon speakers something, it is going to then write me a summary of these speakers about how they are, what they do based on all the information in the proposal. I could have clicked on other steps and see a summary of the floor and all the things that comes with it or the entire exterior or any level that I want. I can dive in and get a written summary. So it's a really, really easy way for me to understand a really complex multi-tier proposal. Forget about me. it allows the client to very easily in a graphic way to follow the different components and dive into whatever section they want without looking at three different proposals with three different PDFs, each and every one of them, 25 pages long and trying to understand how the structure and how it works.

38:37So these are just several different ways on how you can enhance the deliverable of your proposal to make it easier for your client to actually understand what it is that you're offering, which dramatically increases the chances that they're actually going to hire you to do the job. That is it for today. If you enjoyed this episode, please hit the subscribe button and also give us a nice review on either Apple Podcasts or Spotify, wherever it is that you're listening to this podcast. And while you're at it, click on the share button on the corner of your screen and share this with other people that can benefit from this podcast.

39:10I'm sure you know a few people and if you do share it with them, they will appreciate it. I will appreciate it. And you've done a good thing and will feel great about yourself. So why don't do this? It will take you about three seconds to do. That's it for today. And until the next episode, have an amazing rest of your week.

From the publisher

Are you still spending hours writing proposals… only to get ghosted by clients?

It’s time to flip the script. In this solo masterclass, Isar Metis breaks down how business leaders can use AI tools like ChatGPT, Claude, and NotebookLM to craft winning proposals — faster, sharper, and way more aligned with what clients actually want.

You’ll discover how to go from a recorded Zoom call (or in-person meeting) to a fully polished proposal — in minutes, not hours. Plus, you’ll learn how to deal with RFPs (even the messy government ones), write emotionally intelligent cover letters, and use visual tools that make complex proposals client-friendly.

Whether you're pitching services, closing enterprise deals, or trying to impress a board — this episode will transform how you sell.

Learn more about Advance Course (Master the Art of End-to-End AI Automation): https://multiplai.ai/advance-course/ 

Learn more about AI Business Transformation Course: https://multiplai.ai/ai-course/ 

About Leveraging AI

If you’ve enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

More from Leveraging AI

All 330 episodes
238 | Win More Business with Less Effort - How to Write Winning Proposals in 15 Minutes Using AILeveraging AI · 40 min
Listen in VO