There is a huge difference between learning technology as an abstract subject and using it to build a solution that solves a real problem. It was with this idea that the application was born Market Intelligence, a web tool created to support regional studies of competition, opportunities and commercial positioning.
The project was developed as a practical demonstration of the concepts worked on in the courses App Developer e Artificial Intelligence, by My Robot Barra da Tijuca.
The application is published in:
https://gray-glacier-0dc52610f.7.azurestaticapps.net/
And the source code is available at:
https://github.com/ussantos/inteligencia-de-mercado
The idea: transform business questions into a digital experience
The initial question was simple:
How can you help someone better understand a region before opening, positioning or adjusting a business?
In practice, this question involves several others:
- Who are the close competitors?
- What types of alternatives does the customer find in the region?
- Does the chosen address seem to make sense?
- Is the operating radius small, wide or saturated?
- Are there barriers, opportunities or points of attention?
- What commercial actions could be prioritized?
Answering this manually requires research, map reading, data comparison, and some strategic thinking. The application's proposal was to transform this process into a guided flow: the user enters the field of activity, a reference location, the analysis radius and, if they already have an existing company, they can also enter the CNPJ and customer zip codes.
From there, the system organizes public data, consults external APIs, creates a regional map, identifies competitors and generates a report with recommendations.
What came from the App Developer course
The App Developer course works on an essential idea: applications are not just pretty screens. An app needs to have logic, flow, interface, data, tests, corrections and continuous improvement.
This project followed exactly this logic.
Before writing code, it was necessary to think about the user journey. The first version was more complex: it required a login, placed the CNPJ as the central point and showed a lot of technical information. Over time, the experience was refined to become simpler to use:
- login is no longer mandatory;
- the CNPJ became optional;
- the user can analyze an existing company or study a new business;
- the main field became the branch of activity;
- customer zip codes only appear when they make sense;
- the report became more direct and easier to print.
This process represents an important development practice: build, test, listen to the real behavior of the application and improve.
On the technical side, the project uses Next.js, React, TypeScript e Tailwind CSS to create the interface. The application also has API routes, form validations, reusable components, an interactive map, graphs and a report prepared for printing in PDF.
In other words, it is not just a static page. It is a complete web application, with data entry, processing, integration with external services and structured output for the user.
What came from the Artificial Intelligence course
The My Robot Artificial Intelligence course works on concepts such as generative AI, Python, APIs, chatbots, agents, data, context and web systems. The application of market intelligence uses this same mentality: AI does not appear as an ornament, but as part of a larger flow.
An important point is that the system does not just depend on a loose question asked to an AI. Before calling any model, the application structures the context:
- branch of activity informed by the user;
- reference address or zip code;
- analysis radius;
- types of competitors chosen;
- competitors found in public sources;
- observed barriers and opportunities;
- Customer zip codes, when sent;
- limitations of the APIs used.
With this context, AI can help improve recommendations, action plans, positioning and strategic reading. The goal is to generate something more useful than a generic answer.
This is a central lesson for any AI project: the quality of the response depends heavily on the quality of the context. Artificial intelligence works best when given organized data, a well-defined task, and clear boundaries.
Public data, APIs and automation
The project also shows how modern applications are often born from the combination of multiple sources and services.
Among the resources used are:
- Google Places API, to search for competitors and relevant locations;
- Google Maps JavaScript API, to display the map of the region;
- LocationIQ and geocoding fallback, to transform addresses and zip codes into coordinates;
- OpenRouteService, when configured, for support on routes and distances;
- PostgreSQL with Prisma, for cache, history and sharing;
- Azure Blob Storage, for temporary file upload;
- Azure Static Web Apps, for publishing the application;
- OpenAI API, optionally, to enrich recommendations.
The user doesn't need to see all this complexity. For him, the experience should seem simple: fill in data, start the analysis and receive a report.
But behind this there is automation: validation, normalization, external consultation, error handling, data organization, generating recommendations and preparing a layout for printing.
LGPD and data care
Even though it is an educational project, the application was designed with attention to privacy.
Sending customer zip codes is optional and only appears when the user informs that they already have a company. The guideline is to only send zip codes, without names, telephone numbers, emails, CPF or any unnecessary personal data. When a spreadsheet contains other columns, the system must only consider the information necessary for the regional analysis.
Additionally, files submitted for temporary processing may be removed from storage after analysis. This decision reinforces an important idea: useful technology also needs to be responsible.
The report: less noise, more interpretation
An important part of the project was realizing that a good report is not the report with the most information. It is the report that helps the user decide.
Therefore, the application output was adjusted to reduce redundancies and better explain each section. The report presents:
- summary of the region analyzed;
- relevant competitors around;
- reading barriers and opportunities;
- practical recommendations;
- map and graphs;
- Business Strategic Canvas;
- action plan;
- observations on public sources and limitations.
When the user does not send customer zip codes, the application does not force an affinity analysis or customer neighborhoods. In this case, it focuses on what really exists: competitors and market signals around the informed location.
This choice is important because it avoids confusing competition with a customer base. Data should only be used when it represents what it claims to represent.
The biggest learning
This project shows that learning development and AI does not mean memorizing tools. It means understanding how to transform a need into a solution.
On the development side, it was necessary to think about interface, validation, responsiveness, backend, database, deployment, printing and maintenance.
On the artificial intelligence side, it was necessary to think about context, data quality, prompts, APIs, limits, cost and responsibility.
In the end, the tool is a bridge between the two courses:
- o App Developer appears in the construction of experience, in the logic of the system and in the delivery of a functional application;
- the course of Artificial Intelligence It shows up in the use of data, APIs, context, and AI-assisted recommendations.
The result is an educational application, open for study, adaptation and improvement.
Project links
- Published application: https://gray-glacier-0dc52610f.7.azurestaticapps.net/
- Repository in GitHub: https://github.com/ussantos/inteligencia-de-mercado
- App Developer Course: https://www.myrobotbarra.com.br/app-developer.html
- Artificial Intelligence Course: https://www.myrobotbarra.com.br/inteligencia-artificial.html
Final observation
This project has an educational character. It does not replace professional analysis, legal, accounting, labor, tax, technical or data protection validation. Any organization wishing to use, adapt, or deploy the application must review the code, configurations, data sources, API costs, and responsibilities involved.
As a learning exercise, however, it shows something powerful: when application development and artificial intelligence are taught in a practical way, students stop being just users of technology and start to understand how digital solutions are designed, built, tested and published.
Courses that appear in this project
This example connects app development, AI, data, APIs, and web publishing. Therefore, it talks directly to two tracks from My Robot Barra da Tijuca.
APP Developer
For young people who want to create applications, interfaces, logic in Python and digital solutions with beginning, middle and delivery.
Discover the course
Artificial Intelligence
For teens to learn generative AI, data, APIs, context, and creating solutions with critical thinking.
Discover the courseAccess the demo and code
The project is published for testing and can also be studied in the repository at GitHub.
Do you want to see technology applied like this?
At My Robot Barra da Tijuca, courses like APP Developer and Artificial Intelligence help teenagers understand how digital solutions are designed, built, tested and published.