How AI assistant you can make 10 times more productive: a step-by-step instructions

Let’s be honest: AI assistants are no longer just for managers. You are the secret weapon that you can make 10 times more productive than everyone else in your company.

I have tested ai assistant in the past few months. I found that you can change your work with the right setup. During this process I developed two groundbreaking changes AI assistantwith a strategic project work and another to the treatment of assistance tasks of managers.

In the following I will dissolve my input requests and then compare the main actors -Google Gemini, Claude and Chatgpt -so that you can select the perfect AI assistant for your specific needs. Let us immerse yourself.

Everything starts with their folders.

Let’s start with something that nobody talks about, but is absolutely critical: your folder structure. It is not noticeable. With this secular step, however, you save time to set up AI assistants.

In my role at Drift Kings Media, I manage three pods. I only have two folders for each pod. A folder has my strategy documents such as street maps and memos. The other contains all of my meeting transcript-Jeden zooman call, every brainstorming session and every check-in. To automate this, I use Zaps. Every time a meeting ends, the transcript is only in the right folder.

This sorting system is not just about cleaning up and organizing. I actually build the basis for my AI assistants to thrive. My strategy folder trains my project assistant. My transcripts will be my executive assistant.

The basis for the knowledge of your AI is everything structured and unstructured. So if you do your folders correctly, everything else will follow.

Building an AI -Executive assistant

Let’s take this information now and create our first AI assistant. This GPT helps me to prepare for meetings and summarize calls so that I can remain organized. In the following I will build my assistant with Google Gemini.

Gemini has a chat function as well as “gemstones” or Google’s equivalent to custom GPTs. For this process, I will start with a new jewel that the assistant of Executive will be for my scaled POD.

Next I add my PRomption to the instruction section. I want this to analyze my meetings so that I know the action elements that I have to complete after my calls. Here is what I wrote:

• You are my executive assistant for an initiative called Scaled Selling. I want you to:

• Analyze meeting transcripts.
• Identify important decisions and action elements.
• Track periods and possessions.
• Flag with high priority articles.

I also give him the exact starting format that I want. For the meeting analysis, this includes:

• A summary of the meeting (2-3 sentences).
• Key decisions.
• Objects of action (with owner and deadline).
• Flags with high priority.
• Proposed follow-ups.

With my inputs I can upload additional knowledge to train my GPT. Here my folder structure comes into play. I just go to drive and just grab my meeting transcript folder for this pod.

As soon as my assistant has access to all my transcripts, my AI -Executive assistant is ready to go. When I have a new meeting, I download the new transcript. Then my GPT gives me a summary of the meeting with objects of action and deadlines.

Building an AI project assistant

Next I create a project assistant from the information in my strategic documentation folder.

Just like I did for the assistant of the business area, I click on “Edit”, give it a name and add the input requests to the instructions. However, I need more robust instructions because this GPT takes over over higher strategy work. So my request is much longer.

I want my assistant to carry out a few functions. First of all, I would like to upload a new document with my project. If my assistant receives this information, I would like to check the information in the document, summarize important points, determine and track the responsibilities of the stakeholders, how resources are assigned.

From there I would like to create a menu with services from which I can choose. Everything I have to do is to put on a number and my GPT should give me the required edition. My key services include:

• A status analysis that tells me how far we are in the project.
• Risk reviews that help me plan potential blockers.
• Tactical generation that emphasizes and recommends various components of my project how I can make everyone successful.
• Strategic recommendations so that I can prioritize and improve my approach for the project.
• Documentation generation to help me create status reports and progress summaries.

I will specify my input request below. However, I recommend that you adjust these instructions so that you meet your requirements.

The command prompt

You are a AI project assistant with specialist knowledge in relation to project management, risk analysis and organizational dynamics. After analyzing the project documentation provided, you present a menu with available services and immediately carry out the selected service.

First documentation check

Analyze all the project documentation provided, including, but not limited to:

• Project charter and goals
• Strategic planning documents
• Team structure and responsibilities
• Timeline and milestone tracking
• Budget and resource assignment
• Risk reviews
• Technical specifications
• Communication protocols of stakeholders

Menu available

1. Project status analysis

• Comprehensive review of the current project state
• Progress evaluation against goals
• Identification of critical bottlenecks
• Analysis of the milestone rates
• Review of the resource utilization

2. Risk assessment

• Identification of current and potential risks
• Effect and probability analysis
• Detailed recommendations for reduction
• Early warning indicators
• Emergency planning proposals

3. Tactical generation

• Identification of key components to make the project successful
• Deep analysis and thoughtfulness of tactics to make this component a success
• Step-by-step instructions for implementing tactics
• Add important obstacles to make tactics a success

4 .. Strategic recommendations

• Action elements with a high impact
• Priority initiatives identification
• Specific improvement strategies
• Success metrics definition
• Implementation approaches

5. Documentation generation

• Status report creation
• Action article tracking
• Decision protocol maintenance
• Progress summary

Now I only have to upload my strategic Docs folder from Google Drive to my gem’s knowledge section. As soon as this is done, the project assistant analyzes my documents and creates expenses based on the above -mentioned input requests.

With my new project assistant

Now it’s time to test my project assistant. First of all, I will ask me to give me a summary of my project status. Then my GPT provides important successes and gives an overall picture.

From there I can ask for strategic recommendations or a one -sided summary for my management team. My assistant will give me strategic recommendations that I have not thought about, or the most important successes based on presentations that we have made.

Instead of spending hours that sort manually through meeting notes and persecution of action articles, my assistant automatically treats all of this. It analyzes entire folders of documents in seconds, creates patterns over conversations that would take days to find and create perfectly formatted reports with my exact templates.

The real game-change is not only the time that is stored in administrative tasks, but that the AI ​​proactively identify risks, blind spots can connect blind spots and the points between non-connected teams. This is not just an incremental improvement; It is a productivity that liquidates you before your competition.

Compare Google Gemini, Claude and Chatgpt

Now that I showed you the steps in Gemini, I would like to guide you through the same process in Claude and Chatgpt so that you can see the strengths of the individual.

In Chatgpt I created custom GPTs that were trained on the same data as my gemstones. Importing meetings with transcripts and strategic documents into the platform is just as manually as Gemini. I have to click on the “Add attachment” Message “Message GPT Builder”. Nothing can be imported automatically.

I like how Chatgpt creates a clickable menu with my input requests. After I have uploaded a document, I can tell the assistant what to do with one click. However, I found the assistant function of the Executive Assistant between Gemini and Chatgpt.

What is really noticeable is the project assistant. I can add several strategic documents and carry out an analysis. I was surprised at how well the GPT gave recommendations. I felt like I was working with a strategic partner who could help me refine my thinking. Gemini was able to perform similarly detailed outputs, but I had to delete it with further follow-up questions.

I also tested Claude to see how his AI assistants cut off. I found this system more difficult to work. On the one hand, the context window is very small compared to Gemini and Chatgpt.

I was still able to add to commitment requests and connect folder. With Claude, however, you have to add files one after the other, which feels less intuitive. Claude also has inputs for the formatting of inputs. For example, I cannot import PowerPoint foils. This is not a problem with the other two platforms.

However, the project assistant’s recommendations are great. In fact, I found the quality of writing better than Gemini. The establishment of the assistant only requires in this interface.

Ultimately, Chatgpt wins for his project assistant edition and the intuitive interface. But I still use Gemini every day because it is connected to the Google Suite of products that I use every day.

Increase productivity with AI assistant

The head here is not the output, but the process, but the process of how I got there: structure documents in two folders, create two AI assistants for each project and create a menu with options that delivers my desired edition. For me, the integration of this daily work has a major impact.

I recommend using your preferred platform to create your AI assistant. From there you can use other AI tools as companions to improve your results. Sometimes I take the edition of Gemini, put it in Claude and ask you to rewrite it for another audience like my management team. Take your time to experiment, mix your AI tools and correspond until you have a river that works.

In order to learn more about the increase in productivity through the construction of AI assistants Complete episode from Marketing against the grain: