Working as a Lead Product Designer and Manager • Building a whole app experience • Having 100k downloads in 8 months since launch
dinAI was my first opportunity to work across every field of Product Design and Management in a fast-paced environment. Our team of 7 people could build, test, develop, and make this finance app available at the app stores within 6 months. After 8 months since the official launch, we've helped more than 100k users generate personalized investment recommendations with the help of our financial team knowledge and Artificial Intelligence implementations.
Introduction
A high performing team of 7 people developing a finance app that recommend investments in 6 months using agile methodologies, artificial intelligence and a lot of research. dinAI is available in the app stores, don't miss out.
dinAI began as a partnership between the popular finance YouTube channel Gemeos Investem, which has over 2 million subscribers, and a tech team composed of 4 people. I was brought on board as the Product Designer and Manager, serving as the sole designer in the company. The project initially aimed to help users manage their income and expenses, but after extensive research and exploration, it evolved into a product focused on recommending investments.
Our first meetings took place in April 2023. By late November 2023, we had already begun testing through Apple’s TestFlight for debugging. By May 2024, dinAI went live on both the Google Play and Apple App Store. By the end of 2024—approximately eight months after launching—we had surpassed 50,000 monthly active users, accumulated over 100,000 downloads, and achieved a 20% Day 7 retention rate average.
Empathizing
We conducted wide-ranging research—surveying over 1,000 respondents, interviewing 15+ people, and benchmarking against major finance platforms and top design apps. We discovered that 60% of Brazilians lack investing knowledge, 22% still rely on low-yield savings, and many distrust financial advisors. These findings helped us pinpoint our core problem, define our target audience, and identify gaps in the market.
To better understand the problem we aimed to solve, we conducted two quantitative research studies with over 1,000 responses, plus more than 15 qualitative interviews. We also performed extensive market benchmarking against players like XP, Rico, Investidor Sardinha, and Investidor 10, as well as visual feature benchmarking from apps such as Spotify, Airbnb, and NuBank (covering both successful finance applications and the best design benchmarks available). In addition, we drew insights from well-known research reports like “Raio X do Investidor” by ANBIMA and, of course, some data available at IBGE, which is one of the main independent research institutes in Brazil.
It is important to say that each one of these researches aimed for different objectives: to understand the demographics of our users, to understand the potential problems people were facing in order to begin investing, what their motives to start investing were, on what our benchmarks were performing well and on what they were performing badly and so on… In that way, we would be able to make good decisions in our product development process by structuring our data and findings through some well-known frameworks (such as SWOT, and etc.)
Key findings from the market and quantitative researches:
• 60% of the Brazilian population don't have any knowledge about how to invest;
• 22% of the Brazilian population invest in the worst possible investments (in regards to profitability and risk) named "Poupança";
• 33% of the Brazilian population look out to invest in order to buy their own home;
Key findings from the qualitative interviews:
• Most of the interviewed people that already invest, do not trust their investments advisor because of conflicts of interest;
• They also feel afraid when investing due to the lack of information available (and risks involved);
• Most of the interviewed people don't know exactly how their investments are performing;
Armed with all this information, we gained a clear picture of the biggest challenges faced by individuals looking to begin investing (and who were these individuals!)
With that understanding, our team set out to define both the exact problem we wanted to tackle and the audience we aimed to serve. Throughout this discovery phase, we uncovered gaps in the market, identified whom we could best help, and developed a strong overview of the industry landscape.
Defining
Over 70% of users struggle with knowing where to invest and tracking performance. Our research revealed that lack of trust in recommendations, limited financial education, and short-term thinking were key obstacles. By defining our core persona (a young Brazilian new to investing) and clarifying the main challenge—helping users know where to invest—we set the stage for ideation and solution testing.
During our discovery process, we found that determining where to invest and how investments are performing (with over 70% of the respondents) are one of the most significant challenges users face when starting to invest. Naturally, this is a broad topic, which we explored more thoroughly through interviews with potential users. Along with defining the problem at this stage, we also developed our persona through research: a young (18-25 yo), Brazilian male who is just beginning to learn about investing.
From our findings, we identified that people often struggle with deciding where to invest due to a lack of confidence in recommendations from various channels, limited understanding of the investment market (financial education), and an absence of a long-term perspective on the benefits of investing. With a clear problem statement in hand—helping people know where to invest—we were ready to brainstorm and generate solutions (and to test them out with our potential users!).
Ideating
We rapidly brainstormed solutions on FigJam, using techniques like “worst possible ideas” and “rapid ideation.” Within a month, we tested multiple concepts—covering features, branding, and communication—to address our core needs: financial education, investment recommendations, and tracking. My role involved translating ideas into rough sketches and sharing prototypes with both the internal team and interviewed users, ensuring we selected the most viable path before proceeding to user testing and production.
Using FigJam to quickly gather and explore different ideas to solve the above-mentioned problem statement, we facilitated rapid brainstorming cycles employing techniques like “worst possible ideas" and "rapid ideation". Within one month, we tested and presented numerous concepts to the team, allowing us to evaluate and select the most promising direction. At this stage, we looked beyond product features to include branding considerations as well. Equipped with a deep understanding of our persona and product goals—based on our defined problem—we were also able to determine how we wanted to communicate with our users.
Our product needed to address financial education, provide investment recommendations, and offer investment tracking. This phase was crucial. The team generated multiple potential solutions, and my primary role as both product manager and designer was to translate these ideas into rough sketches. By sharing these early prototypes with our internal team and some of our interviewed participants, we could gauge whether our concepts effectively solved our persona’s challenges. Once we identified the most promising approach, we moved on to user testing before finalizing and moving into production.
Fast prototyping
We rapidly built and tested streamlined user flows using Figma and componentization, drawing insights from industry leaders like Cameron Worboys (Wise) and Katie Dill (Stripe). By leveraging resources such as ConFIG videos and design system best practices, we evolved from low-fidelity prototypes to a robust, brand-aligned design system for the dinAI app.
My approach to prototyping centered on creating streamlined user flows and reusable components, allowing us to validate concepts quickly and confirm we were delivering real value. To refine our techniques, I turned to ConFIG videos on YouTube and insights from industry experts like Cameron Worboys (Wise), Andy Allen (!Boring), and Katie Dill (Stripe, Airbnb), which highlighted best practices for rapid prototyping and design system development. By leveraging these resources, we were able to build and iterate low-fidelity prototypes in Figma, then progressively refine them through user feedback loops. This fast, iterative process laid the groundwork for what would eventually become the dinAI design system.
Following the Nielsen Norman Group’s recommended best practices for the Prototyping stage, we conducted multiple rounds of testing and iteration to ensure our concepts effectively resonated with users. By gathering feedback early and often, we could pivot quickly when ideas fell short, reduce risk, and confirm that each design element delivered meaningful value. This user-centric, iterative approach helped us maintain alignment with our core objectives while still remaining flexible and open to insights that emerged throughout the Prototyping phase.
Testing
We tested the prototype with our team and a group of about 50 potential users via WhatsApp, gathering and categorizing feedback to prioritize weekly pain points using sprint methodologies. We further refined our concept through a Figma prototype and user interviews, working closely with our target audience. Our main goal was to solve the user’s defined problem quickly, validate the idea in a small group, iterate based on feedback, and then launch fast to test at scale and explore monetization potential.
During the testing phase, we introduced the prototype to our team and a group of approximately 50 potential users via a WhatsApp community, gathering direct feedback on usability and perceived value. We meticulously structured and categorized these insights, identifying recurring pain points that we tackled on a weekly basis using sprint methodologies. To further validate our concepts, we utilized a Figma prototype and conducted user interviews, collecting contact information via online forms so we could maintain ongoing communication. This real-world input was instrumental in ensuring our solution effectively addressed user needs before moving toward broader deployment. Our core objective at this stage was to validate the concept with a small audience, iterate rapidly based on feedback, and confirm we were on the right track in solving the identified problem.
Following the Design Thinking Methodology best practices for the Testing phase, we adopted an iterative approach—running multiple rounds of short, focused tests to capture qualitative and quantitative insights. By observing actual user behavior and refining design elements promptly, we minimized development risks and stayed aligned with user expectations. This user-centered, test-early-and-often strategy enabled us to make informed decisions and quickly prioritize improvements that had the most significant impact on the overall user experience.
Continuous development
After launching dinAI, we shifted into continuous development driven by daily analytics checks via Amplitude, user feedback from multiple channels, and agile best practices. This approach led to over 100,000 downloads, a 25% Day 7 retention rate, and more than 50,000 active users. We introduced more than a dozen new features, refined our product roadmap with structured feedback, and prepared future offerings—like a mentoria program—to sustain long-term growth and value.
With dinAI live, we shifted into a continuous development phase that emphasized structured data analysis and rapid iteration. We created in-app events to gain deeper insights into actual user behavior—recognizing that what users say they do can differ from their actions in the app. We also gathered feedback from multiple channels—app store reviews, in-app feedback prompts, online forms, and dedicated WhatsApp groups—to guide dinAI’s evolving priorities in each sprint.
After launch, we persisted in testing and refining our solutions, leveraging daily analytics checks in Amplitude to monitor user engagement and quickly spot opportunities for improvement. Despite operating in a dynamic environment, dinAI surpassed 100,000 downloads, maintained a 25% Day 7 retention rate, and reached over 50,000 active users (paid and free) within ten months. Throughout this period, we rolled out more than a dozen new features—some quite complex—while adhering to digital product management best practices, including Scrum ceremonies and insights from industry experts like Marty Cagan.
From a product operations perspective, our continuous development efforts focused on several key areas. First, Feature Deliveries ensured we continually released new capabilities to meet user needs. Second, Data Analysis relied on daily tracking via Amplitude, allowing us to measure user behavior and feature success in near real-time. Third, Feedback Categorization provided a systematic way to inform our product roadmap and weekly sprint goals. Finally, our Product Strategy included plans for future offerings—like a dedicated mentoria program—aiming to further extend dinAI’s value proposition and stay aligned with user expectations.