About this project
Hey! I'm Florian, the creator and developer of QuickDraw.
QuickDraw is a personal project born from a very real need: performing reliable calculations in the middle of a Pokémon Champions battle. Faced with the complexity and lack of ergonomics of existing tools, I wanted to create something more in line with the values of the game itself : accessibility, simplicity, modernity. It all started with a damage calculator.
The rest built itself naturally: a team builder to load your Pokémon in one click, meta statistics from tournaments to simplify input, and so on. What was meant to be a single tool grew into a full competitive suite.
I work on this project alone, driven entirely by my passion for development and my love of working with data; and playing Pokémon Champions, of course. App development is my profession, and thanks to the recent rise of generative AI and Claude Code, I get to bring this project to life in my free time. I design, shape, and develop every feature with care.
I hope you enjoy using it as much as I enjoy building it!
AI Usage Transparency
I've been a professional software developer for almost 10 years. I also hold a PhD, which gave me a lasting passion for data analysis and extracting meaningful insights from numbers. And most importantly I've loved Pokémon for as long as I can remember.
When Champions launched, it brought all three together in a way no game had before. I wanted to build something meaningful around it; but meaningful projects take time, and time is the one thing a solo developer working outside office hours doesn't have.
AI made this project possible. Not by replacing the work, but by compressing it. The same way a composer today can bring a musical idea to life using software tools that would have taken an orchestra and months of arrangement work, AI allowed me to express something I genuinely care about.
QuickDraw is the result of that. Everything that follows explains precisely what that means in practice.
What AI does
AI is used at two stages of development. First, as a sounding board during the design phase: helping refine ideas, anticipate edge cases, and stress-test assumptions before a single line of code is written. Second, as a code generator : once a detailed specification is ready, AI produces a first implementation based on precise instructions, including which components to create or reuse to maintain consistency across the project.
For debugging, AI is occasionally used to identify the root cause of complex issues spanning multiple files, producing a preliminary report on causes and recommended fixes before any correction is applied.
What I do
Every feature starts with a design and layout concept, a functional specification, and a set of explicit development constraints. The AI output is then tested, reviewed, and rarely correct on the first pass. Gaps in the specification lead to extrapolations that need to be identified, documented, and corrected one by one, with precise instructions for each adjustment.
Structural bugs are fixed directly without AI involvement. Calculations, be it damage results, meta stats or whatever equation being used (the core of QuickDraw's reliability) are verified manually. This is not something that can be extrapolated or approximated: every variable in the equation is checked personally before any feature goes live.
The final stage of every feature is a full test cycle covering for example multilingual behaviour, interface responsiveness, and API integrity. Only then is it deployed.
Why AI
QuickDraw would not exist without AI. Not because the technical challenge was beyond reach, but because building a project of this scope (a full competitive toolkit with a polished interface, multilingual support, real-time calculations, and regularly updated data) requires hundreds of hours that a solo developer working outside office hours simply doesn't have.
AI compressed that time. It didn't replace the decisions, the design, the validation, or the expertise : it made it possible to act on them within the constraints of real life. The tools just removed the bottleneck.
A final note
There are no fixed rules about what AI can or cannot touch in this project. The balance is determined by my own judgement, not by principle. AI is a collaborator with specific strengths and limitations; knowing both is what allows me to use it effectively. Some problems call for its involvement, others call for direct expertise. In the end, the decision is always mine.
What doesn't change is the standard: every feature is tested, every calculation is verified, and nothing ships until it meets the bar I've set for a tool I use myself.
Upcoming features
QuickDraw is constantly evolving. Here are some of the features I'm working on:
In-depth tournament analysis : explore tournament data in detail to better understand metagame trends
Pokémon suggestions : recommendations to complete your team's offensive coverage
Singles format (3v3) : data and tools integration for the Singles format
Support me
Want to support this project? You can do so by offering me one of the many daily coffees required to push it even further!
Buy me a coffee