SUPER LAUNCH AI AGENT MANAGER TRAINING 2.0

TAKE ADVANTAGE OF THE SPECIAL OFFER

Hours
Minutes
Seconds

How to plan your application | The 7 pre-development phases

application planning

Learn once and for all how to plan your application development project, which steps you cannot miss in your planning and, on top of that, learn about some tools that can help in this process.

One of the biggest mistakes that many developers make when starting a project is jumping straight into development, be it with nocode tools or not, before even planning the application.

Some do this because they think they know everything that will be needed and developed in this project and they think that this will be more productive,

Others do this due to a lack of knowledge in best softwares development practices.

You can be sure that these projects will either have holes or will take much longer than necessary to complete, all due to the lack of good initial project planning.

This is not only valid for projects with external clients, but it is also valid for the development of your own personal projects or side projects.

In this video I want to show you the 7 phases of the NoCode StartUp method for carrying out application planning.

These are extremely important phases that can and should be carried out even before we enter the development stage within the tool.

There is a lot of planning that can be carried out, even before we want to open the platform that we are going to use to develop our application. Be with Bubble, FlutterFlow, WeWeb, AppGyver, any nocode tool or even (and especially) with code.

We divided our planning into 7 phases, which I will present here for you now, let's show this visually to you.

In the end, I still want to present to you some tool tips that can help us a lot in some of these phases.

Phase 0 – General Application Concept

Phase zero is the general conceptualization phase of the application, that is, here we will define exactly what our app is, what its objective is, what it does, etc…

This phase is extremely important as it will be the basis for all the others.

Everything comes from this conceptualization, which is why this mapping must be extremely well done and aligned with the client or anyone involved in the project.

In the end, this ends up practically becoming a scope of the project, what will be done and what is expected.

Here, as an example, I brought some of the points that can be raised at this stage of planning the application:

App concept: in this case we are talking about a multi-company project management app.

We can collect requirements, or basically functions of our app. Listing what we expect the app to do.

(We have some content about requirements gathering, which you can see here on our blog or on our YouTube channel.)

We can list the pages that our app will have, type of users, user permissions and so on.

Phase 0 - Application Planning
How to plan your application | The 7 pre-development phases 8

I believe you understand the importance of this step, right? Our entire app will be based on what we collect and write down here.

Phase 1 – Inspirations

Now that we know the objective of our application and all its requirements, we can look for inspiration.

Look for applications that do something similar to ours, to be inspired mainly by Usability.

In design there is a law called Jakob's Law – which says:

“People spend most of their time on other websites and prefer their website to work in the same way as all the other websites they already know”

In other words, users expect their website, app, system to have usability similar to the other apps that exist.

Important point: the idea here is to inspire us, NOT COPY.

Here in our example, we are creating a project manager and we already know the pages we are going to develop. This way we can look for inspiration in similar applications.

Phase 1 - Application Planning
How to plan your application | The 7 pre-development phases 9

We brought here some inspiration for registration and login flows.

Some inspirations on how some project management systems like ClickUp or Asana show their projects to users and so on.

This way we begin to have an idea of how the market already does what we are trying to do, we can be inspired and, on top of that, improve the UX.

Now that we know the pages we are going to develop, what our app needs to do and we also have some inspiration, we can start designing our Wireframes

Phase 2 – Wireframing

Wireframes are nothing more than a low-fidelity design of our application, that is, here we will not focus on placing colors, visual identity, details, but rather we will focus on creating a first sketch of the elements that will be present in our app, in the their disposition and the experience we hope to pass on to our users.

This step is crucial, because with it we can start to visualize the face of our application and we can also get quick and objective feedback from those involved in the project.

With Wireframes we can remove distractions such as colors, design and focus on collecting feedback exactly about what matters at that moment, layout and usability.

Phase 2 - Application Planning
How to plan your application | The 7 pre-development phases 10

Here in our example, we can already see what our login flow will look like, how our dashboard will be laid out and so on.

Phase 3 – User Flows

This phase is very common to be carried out in parallel to wireframing and aims to document and detail all the flow of actions that each user can perform on the screens in question.

We detail all actions, permissions and restrictions considering each type of user for each screen or page.

Phase 3 - Application Planning
How to plan your application | The 7 pre-development phases 11

In our example here, on our login page, we detail the flows:

  • New Registrations
  • Login

Phase 4 – Data Modeling

I believe that this phase does not require such an important comment.

Data modeling is the heart of any application and must be done before we think about jumping into any platform to start development. This is what differentiates apps that will work well when there are more users than apps that will not.

Without good data modeling, applications are destined to cause major problems in the future. Failures in modeling can lead to slowdowns, a drop in performance and in some cases the solution will be a complete refactoring of the app.

Since data is the heart, you end up building your app based on data modeling. The logics end up being designed according to what was designed in this modeling. That's why it's important to invest a lot of time in this modeling before even thinking about using the tool.

Phase 4 - Application Planning
How to plan your application | The 7 pre-development phases 12

Here in our example we can see which tables will be needed in our app, which fields we will have in each table and how they relate to each other.

Here our objective is not to teach how to do data modeling, however we have two free courses on YouTube on the topic, one on relational data modeling and another about non-relational data modeling, I highly recommend you watch it, I'll leave the video cards listed here.

Phase 5 – Security

This here is the most underestimated phase by all novice users and even some users with years of experience, it is a bureaucratic step, but extremely necessary and should also ideally be thought of before we start developing our apps, as there may be cases where We need to remodel some areas of our database to be able to implement 100% the expected security in our app.

In practice, this implementation varies from tool to tool, but at a conceptual level, the idea is the same.

We need to think about the fields in our database and basically say which fields can be seen by which type of user.

By doing this mapping, then we just need to implement it in our systems.

Phase 5 - Application Planning
How to plan your application | The 7 pre-development phases 13

In our example, I passed Data Type by Data type and implemented the necessary rules to ensure that only those who can actually see the data are the only ones with access to it.

Phase 6 – Visual Identity + High Fidelity Prototyping

Now it's finally time to think about the design of our application.

At this stage we define the entire color palette of the project, default styles, fonts, etc… And we implement this in our application, based on what we have already built in our wireframes.

Furthermore, at this point we can choose to prototype our application in a tool like Figma, for example, and give a sense of life to this design and application.

Phase 6 - Application Planning
How to plan your application | The 7 pre-development phases 14

It is important and interesting to comment here that if we stop to analyze, practically in all the phases mentioned here, we can carry out micro validations with our clients, this way we will advance the project little by little, with the client's approval.

This completely mitigates rework in more advanced stages of development, which take much longer to adjust.

As promised, here are some indications of tools that you may be using to carry out some of these steps:

General Planning:

Inspirations:

Wireframing:

UserFlows

Again folks, all of these steps can and should be done before we even think about opening our NoCode tool.

Of course, as your technical knowledge evolves, this process will become increasingly easier and you can also identify more Phases in this planning process specific to the tool you use.

I know some European and Australian development agencies, which have specific documentation processes and have already thought about the tool they use to develop.

I recommend everyone not to neglect this process, otherwise you will probably understand the reason later.

If you enjoyed this content and are also interested in this topic, within our training we will explore this topic further.

These examples that I gave here in this video are from a complete track that we have in NoCode StartUp Bubble formation, where I detail each of these topics step by step with you and later we build this Project Management application together.

Thank you, big hug and see you next week!

org

Watch our Free MasterClass

Learn how to make money in the AI and NoCode market, creating AI Agents, AI Software and Applications, and AI Automations.

Neto Camarano

Neto specialized in Bubble due to the need to create technologies quickly and cheaply for his startup. Since then, he has been creating systems and automations with AI. At the Bubble Developer Summit 2023, he was listed as one of the greatest Bubble mentors in the world. In December, he was named the largest member of the global NoCode community at the NoCode Awards 2023 and first place in the best application competition organized by Bubble itself. Today, Neto focuses on creating AI Agent solutions and automations using N8N and Open AI.

Also visit our Youtube channel

Learn how to create AI Applications, Agents and Automations without having to code

More Articles from No-Code Start-Up:

Large Language Models (LLMs) have become one of the most talked-about technologies in recent years. Since the meteoric rise of ChatGPT, generative AI-based tools are being explored by entrepreneurs, freelancers, CLT professionals, and tech-curious individuals.

But why is understanding how LLMs work so important in 2025? Even if you don't know how to program, mastering this type of technology can open doors to automation, the creation of digital products, and innovative solutions in various areas.

In this article, we will explain in an accessible way the concept, operation and real applications of LLMs, focusing on those who want to use AI to generate value without relying on code.


What is an LLM

What is an LLM?

LLM stands for Large Language Model. It is a type of artificial intelligence model trained on huge volumes of textual data, capable of understanding, generating and interacting with human language in a natural way. Famous examples include:

  • GPT-4 (OpenAI)
  • Claude (Anthropic)
  • Gemini (Google)
  • Mistral
  • Perplexity IA

These models function as “artificial brains” capable of performing tasks such as:

  • Text generation
  • Automatic translation
  • Sentiment classification
  • Automatic summaries
  • Image generation
  • Automated service

How do LLMs work?

In simple terms, LLMs are built on Transformer neural networks. They are trained to predict the next word in a sentence, based on large contexts.

The more data and parameters (millions or billions), the more powerful and versatile the model becomes.

Read more: Transformers Explained – Hugging Face

Own LLMs vs. API usage: what do you really need?

Building your own LLM requires robust infrastructure, such as vector storage, high-performance GPUs, and data engineering. That's why most professionals opt for use ready-made LLMs via APIs, like those of OpenAI, Anthropic (Claude), Cohere or Google Gemini.

For those who don't program, tools like Make, Bubble, N8N and LangChain allow you to connect these models to workflows, databases, and visual interfaces, all without writing a line of code.

Additionally, technologies such as Weaviate and Pinecone help organize data into vector bases that improve LLM responses in projects that require memory or customization.

The secret is to combine the capabilities of LLMs with good practices in prompt design, automation and orchestration tools — something you learn step by step in AI Agent Manager Training.

Difference between LLM and Generative AI

Although they are related, not all generative AI is an LLM. Generative AI encompasses many different types of models, such as those that create images (e.g. DALL·E), sounds (e.g. OpenAI Jukebox), or code (e.g. GitHub Copilot).

LLMs are specialized in understanding and generating natural language.

For example, while DALL·E can create an image from a text command, such as “a cat surfing on Mars,” ChatGPT, an LLM — can write a story about that same scenario with coherence and creativity.

Examples of practical applications with NoCode

The real revolution in LLMs is the possibility of using them with visual tools, without the need for programming. Here are some examples:

Create a chatbot with Dify

As Dify Course, it is possible to set up an intelligent chatbot connected to an LLM for customer service or user onboarding.

Automate tasks with Make + OpenAI

Node Makeup Course You learn how to connect services like spreadsheets, email, and CRMs to an LLM, automating responses, data entry, and classifications.

Building AI Agents with N8N and OpenAI

O Agents with OpenAI Course teaches how to structure agents that make decisions based on prompts and context, without coding.

Advantages of LLMs for non-technical people

Advantages of LLMs for non-technical people

  • Access cutting-edge AI without having to code
  • Rapid testing of product ideas (MVPs)
  • Personalization of services with high perception of value
  • Optimization of internal processes with automations

LLMs and AI Agents: The Future of Interaction

The next evolutionary step is the combination of LLMs and AI agents. Agents are like “digital employees” that interpret contexts, talk to APIs and make decisions autonomously. If you want to learn how to build your agents with generative AI, the ideal path is AI Agent Manager Training.

We are living in an era where texts, images and videos can now be created by artificial intelligence. But there is one element that is gaining strength as a competitive advantage: the voice.

Whether in podcasts, institutional videos, tutorials or even automated service, the ability to create realistic artificial voice is changing how companies and creators communicate. And in this scenario, the ElevenLabs AI emerges as one of the global protagonists.

What is ElevenLabs
What is ElevenLabs AI? The AI-Powered Voice Revolution 18

What is ElevenLabs?

O ElevenLabs is one of the neural speech synthesizers most advanced on the market. With its technology AI voice cloning and AI-powered text to speech, allows you to create realistic voices in multiple languages, with natural intonation, dynamic pauses and surprising emotional nuances.

Key Features:

  • Human-quality Text to Speech
  • Conversational AI with support for interactive agents
  • Studio for longform audio editing
  • Speech to Text with high accuracy
  • Voice Cloning (Instant or Professional)
  • Sound Effects Generation (Text to Sound Effects)
  • Voice Design and Noise Isolation
  • Voice Library
  • Automatic dubbing in 29 languages
  • Robust API for automations with tools like N8N, Make, Zapier and custom integrations
ElevenLabs FAQ
What is ElevenLabs AI? The AI-Powered Voice Revolution 19

ElevenLabs FAQ

Find out more about the company and news from ElevenLabs directly at official website of ElevenLabs and see the API documentation.

Does ElevenLabs have an API?

Yes, ElevenLabs has a complete API that allows you to integrate speech generation with automated workflows.

With this, it is possible to create applications, service bots, or content tools with automated audio.

Discover the Make Course from NoCode Start Up to learn how to connect the ElevenLabs API with other tools.

Are ElevenLabs voices copyright free?

AI-generated voices can be used commercially, as long as you respect the platform's Terms of Use and do not violate third-party rights by cloning real voices without authorization.

Is it possible to use ElevenLabs for free?

Yes. ElevenLabs offers a free plan with 10,000 credits per month, which can be used to generate up to 10 minutes of premium quality audio or 15 minutes of conversation

This plan includes access to features like Text to Speech, Speech to Text, Studio, Automated Dubbing, API, and even Conversational AI with interactive agents.

Ideal for those who want to test the platform before investing in paid plans.

What is the best alternative to ElevenLabs?

Other options include Descript, Murf.ai and Play.ht. However, ElevenLabs has stood out for its natural voice, advanced audio editing features with AI, API integration and support for multiple languages.

Their paid plans start from US$ 5/month (Starter) with 30 thousand monthly credits, and go up to scalable corporate versions with multiple users and millions of credits.

See all available plans on the ElevenLabs website. However, ElevenLabs has stood out for the naturalness of its voice and the quality of its API.

How does ElevenLabs work?

You submit a text, choose a voice (or clone one), and AI converts that text into realistic audio in seconds. It can be used via the web dashboard or via API for automated workflows.

Examples of using ElevenLabs AI in practice

1. Video and podcast narration

Ideal for creators who want to save time or avoid the costs of professional voiceovers.

2. Automated service with human voice

Turn cold bots into realistic, empathetic voice assistants.

3. Generating tutorials and training with audio

Companies and CLT professionals can create more engaging internal materials.

4. Applications that “talk” to the user

With tools like Bubble, FlutterFlow or WebWeb, it is possible to integrate AI voice into apps.

How to integrate ElevenLabs with NoCode tools

NoCode tools
What is ElevenLabs AI? The AI-Powered Voice Revolution 20

N8N + ElevenLabs API

Allows you to automate voice generation based on dynamic data using visual workflows in N8N. It is ideal for creating processes such as audio customer service responses, automated voice updates, and more.

Discover the N8N Course from NoCode Start Up

OpenAI Agents + ElevenLabs

With the use of AI agents, it is possible to create voice-responsive systems, such as a virtual attendant that speaks to the customer based on a dynamic prompt.

See the Agents with OpenAI Course

Bubble/FlutterFlow + ElevenLabs

Use the API to insert audio into your apps with interaction triggers or dynamic events.

ElevenLabs and NoCode: Open the door to creating experiences with voice AI

AI-generated voice is already a powerful, accessible and potential-rich reality. ElevenLabs is not just a tool, but an engine for creating immersive, automated and more human experiences.

If you want to learn how to integrate these possibilities with NoCode and AI tools, NoCode Start Up has the ideal paths:

The market is changing – fast. Artificial intelligence is no longer a trend, it has become a necessity. Companies are cutting costs, optimizing operations and looking for specialists to implement AI in their daily lives. And this is exactly where the AI profession comes in. AI Manager Course.

NoCode AI Manager Course: What it is, Who it is for and What its Objectives are

THE AI Agent Manager Training It is aimed at anyone who wants to enter the field of artificial intelligence in a practical way, without needing to know how to program.

The main objective is to train professionals capable of delivering automation and real solutions for companies using NoCode tools.

It is ideal for both those who want to offer services and those who want to open their own AI agency.

The training proposal is clear: enable you to bill more than R$14,000 per month working with intelligent solutions — a market that is only growing.

Access the training here

What is Included in the AI Agent Manager Course?

The training is structured in complete knowledge trails, with content organized by theme and level of mastery:

Topics covered:

  • Fundamentals from Zero to Advanced
  • Mastering Automations with AI
  • Creating and Selling AI Agents to Companies
  • Applied Prompt Engineering
  • Using NoCode tools like n8n, Dify, Make, OpenAI and more
  • Integration with WhatsApp, CRMs and payment gateways
  • Ready-to-clone and apply templates

When you sign up, you get:

  • 8 complete formations, including SaaS AI and technical courses from NoCodeStartUp;
  • Access to exclusive community, active and with direct support from instructors;
  • 1 year full access, including the Make paid plan;
  • NoCode Match, a hub of real opportunities in the AI and automation market.

Differences between AI Manager Training and Other Courses

Unlike many generic courses, this training was designed as a complete ecosystem of learning and practical application, with total focus on generating results for the student.

  • 100% classes structured, edited and with step-by-step teaching methods
  • Mentors present, community engaged
  • Trail with beginning, middle and end, organized with teaching methodology
  • Real opportunities and networking with companies and devs
  • Masterclasses with experts who already apply AI in agencies and companies

Take advantage of the offer

What is the cost of the AI Manager Course and Access Conditions?

The promotional value of the training is R$ 1,497 in cash or in up to 12 installments of R$ 157.53 on the credit card.

  • Full access for 12 months
  • 7-day money-back guarantee
  • Updates included at no additional cost

FAQ: Main Questions About the AI Manager Course

Do you need prior knowledge?
No. All content is designed for absolute beginners in AI and NoCode.

How long do I have access?
You will have 1 year of access complete to the platform and updates.

Can I ask for a refund if I don't like it?
Yes. You have 7 days warranty to test and cancel without bureaucracy.

I'm a PRO student. Do I already have access?
Yes. NoCodeStartUp PRO subscribers have unlimited access to the training.

What are the extra costs for tools?
The only initial cost for teaching purposes is US$ 5 to use OpenAI API.

How long will it take for me to see results?
In the first few days, you can create and test your first AI agent.

What do I get when I purchase?
Immediate access to all tracks, bonuses, community, templates, Masterclass and tools.

How to Become an AI Manager?

To become an AI manager, the ideal is to start with training that combines practice and theory in an accessible way.

The NoCodeStartUp course focuses exactly on that, teaching how to create automations with AI without requiring prior programming knowledge.

You'll learn everything from the fundamentals to delivering real solutions using platforms like n8n, Make, Dify, Zapier, and OpenAI.

How Much Does an AI Manager Earn?

According to the market itself and reports from students, an AI manager can earn above R$10 thousand per month, working with consultancies, creating personalized agents or recurring services via intelligent automation.

Which Course Should I Take to Work with AI?

If you are looking for a practical, up-to-date course, with a strong connection to the market and no programming requirements, AI Agent Manager Training is one of the most complete currently.

It combines technical content with real-world application so you can start working quickly.

Invest in Yourself: Become a Professional AI Manager

If you are looking for a way to stand out in the digital market, enter the technology area without having to program and act with something that is growing rapidly, this course is a smart shortcut.

The AI Agent Manager Training provides a clear path, real support and applicable tools for you to work with AI in a professional manner.

It is applied learning with a total focus on solving real problems using artificial intelligence.

Access the AI Manager training now and start building a new future with AI.

NEWSLETTER

Receive exclusive content and news for free

en_USEN
menu arrow

Nocodeflix

menu arrow

Community