business coach app

curated content tips for each user’s work challenge

katalista is a learning app built to help emerging leaders develop soft skills.

as a co-founder, i led all design and product management, in roughly this process:

1. potential user interviews about study habits;

2. synthesis of typical flow patterns, motivations and blockers;

3. group ‘lean inception’ workshop defining proto-personas, features and journeys;

4. wireframes, ui design system, prototypes, usability tests and development handoff;

5. user tests and interviews, opportunity mapping and product/design iterations.


co-founder, product and design


co-founder, product and design






50% adoption; 20% retention on month 2


50% adoption; 20% retention on month 2

product: personalized, contextual recommendations

businesses and employees recognize the importance of continuous learning, but corporate universities and learning tools fail to engage employees effectively.

to address this issue, katalista was launched with the goal of helping employees who may not prioritize learning. the mvp employs three strategies to achieve this:

1. full user autonomy: while offered by companies to their employees, it is entirely up to the user to decide when and what to consume.

2. content aggregation and curation: katalista selects relevant content from top sources and authors, available in multiple formats and mostly free, in portuguese.

3. personalized and contextual ai recommendations: katalista suggests content based on each user's interests, as well as their current work-related challenges.

design: less but better options, in visual cards with key decision-making info

during potential user interviews, participants frequently mentioned that they would like to study more, but lacked the time. however, further questioning revealed that they often spent extended periods of time on social media or streaming platforms.

interviews also revealed diverse format preferences; for example, some users preferred podcasts, while others preferred video.

the app's home screen design built on these insights:

1. to avoid overwhelming users with too many choices, the main screen displays only three recommendations tailored to the user’s stated work challenge.

2. the visual design was referenced in popular streaming and social media platforms, making the learning content more engaging and desirable than traditional office-style learning systems.

3. the content card design is meticulously crafted, presenting essential information such as author/source, format, duration, and price (usually free) in a clear and hierarchical manner.

interviews, snapshots, opportunity trees and story maps

the product discovery methods used went very much along teresa torres’ continuous discovery habits.

we invited tester companies who provided me with weekly schedules for interviews, which i documented as visual snapshots in coda and synthesized as an opportunity solution tree in figjam, mapping all the most promising pains and desires.

new feature sets were first envisioned in story maps, and then different variations were debated and user tested with hi fidelity screens and prototypes.

tokens and components for fast design / development collaboration

all design styles and recurring elements were tokenized and handed off to development as components, streamlining the design and development process for the first version of the product, as well as future iterations.

this modular and structured approach also proved beneficial when designing other products such as the company website, commercial presentations, newsletters, event graphics, and more.

you can check my components handoff specs in figma.

3x more usage and a lot to improve

we launched the first mvp with a small group of testers, interviewed them extensively, and made rapid small improvements, with a focus on streamlining the onboarding (internal prototype here) and post-content experiences.

without any new features, this version drove about three times greater user retention than traditional learning systems.

but it also made evident that there was still a lot to improve.

for the business client, we realized that having proprietary content or custom recommendations was a really important feature – most companies still want to control what content their employees consume – while for the end users we still had a number of hypothesis on how to drive further engagement.

six months in, we tried to tackle some of these issues introducing a major new feature called ‘trilhas’ (journeys). more on that in a future post, but you can check a prototype for now.

find more and download the actual app at