Launching today
Chinilla
Design systems, simulate them and watch where they break
94 followers
Design systems, simulate them and watch where they break
94 followers
Like a flight simulator for systems. Drag components, wire them together, hit play, and watch real traffic flow through your system. See queues fill up, databases choke, and bottlenecks form in real time. Scrub a timeline to inspect any moment. Fix what breaks, run it again. That's how you go from knowing patterns to understanding systems. Export to PNG, Mermaid, or Python for your docs and repos. 7 blocks, 12 behaviors, AI that turns code or prompts into live diagrams. Free in your browser.
Chinilla
Hey product hunt! 🦖
I built Chinilla because I wanted a tool to help me understand how systems work. I tried learning through books and guides, but they're jargon heavy and still didn't give me the intuition of actually seeing and understanding systems visually.
So I built a flight simulator for it. You design a system visually, hit play, and watch traffic flow through it in real time. When your database chokes or your queue overflows, you see it happen. You fix it, run it again, and build the kind of intuition you can only get by watching things break.
What you can do:
Drag and drop 7 universal building blocks
Wire components together and define simulatable metrics and behaviors
12 programmable behaviors: queue, retry, filter, batch, rate limit, circuit breaker, and more
Hit play to simulate and watch real packets flow through your architecture
See bottlenecks, drops, and queue pressure as they happen
Scrub a timeline to inspect any moment frame by frame
Describe a system in plain english or paste code or text (frameworks, papers, etc), Chinilla AI maps it out on canvas
Collapse groups, enter subsystems, explore at any level
Find weaknesses with stability analysis, Monte Carlo with SLO targets, and stress testing
Export to PNG, GIF, SVG, Mermaid, Python code, or a markdown spec
Publish a live interactive link with a MD embed for papers or repos
16 templates to learn by doing (ML pipeline, coffee shop, chat app, rate limiter, and more)
It's free to use in your browser. The demo takes about 60 seconds and doesn't need an account: chinilla.com/demo
Free account gets you 5 cloud projects, full simulation, all export formats. No credit card.
Poured a lot of 💖 and time into this project. I'd love to hear from y'all what you think!
What's next:
Team collaboration (shared canvases, real-time cursors)
Fully fleshed out Duolingo style interactive lessons that teach system design step by step
Tighter simulate-fix loop (inline suggestions when things break)
Integrations (import from Terraform, CloudFormation, Docker Compose)
Better UX polish based on your feedback (seriously, tell me what's rough)
Launch special: LAUNCH50 for 50% off Pro monthly (until April 30)!!
Thanks for reading 😃
This is a sharp idea — most teams only discover their system design flaws after deploying to production and watching things break under real load. We went through exactly this scaling our internal automation infrastructure. The simulation-first approach could save a lot of painful debugging. Curious about the learning curve for non-engineering founders who need to spec out systems but aren't deeply technical on the architecture side.
Chinilla
@thekrew Great question! Bridging that gap is a big focus. A few things I built specifically for non-technical users: the building blocks are scoped to just 7 universal components (instead of throwing unlimited options at you) to reduce cognitive overload. There's a guided wizard that generates a starting architecture from a plain English description. Interactive Duolingo-style lessons (currently in beta and getting fleshed out) that teach concepts step by step. Full docs with examples. And the AI assistant is designed to teach as it helps, explaining what it's doing and why.
For example, one of the starter templates is a coffee shop workflow. Even without technical knowledge, you can drag and drop components to sketch out any process, then use the AI to refine it and wire up behaviors for you. I've personally used it to design non-technical "systems" too, like a plan to drink fewer energy drinks (LOL), exercise routines, and even a simple org chart.
Here's the energy drink one I made by just asking the AI "can you help me design a system to drink less energy drinks?": https://chinilla.com/share/mbxdgntgwpm2 (designs are publishable through a toggle with shareable links from the user dashboard and file icon in the IDE!)
The 7 blocks also aren't tied to specific behaviors either. They're there to help you mentally map out what you want to do while keeping the flow standardized. You can use any component, rename it to whatever makes sense to you, wire it to behave however you want, and the AI will still pick up on what you're trying to do. Would love to hear how it works for your use case! 😃
My first thought was that this seems extremely useful for games like satisfactory and factorio! Great work.
Chinilla
@matt_doneai Thank you! 😃 Actually, Factorio was an inspiration for this while I was sketching out the idea haha. That feeling of watching your factory choke because you didn't plan throughput right? Same energy here except with diagrams.
To say I'm surprised is to say nothing! Cool idea!
P.s. I'm the only one who didn't get this link to open
Chinilla
@julia_zakharova2 Thank you! 😃Ah! About the link, slight typo! haha
Thank you for catching that. fixing now!
@chinilla It happens)
Chinilla
@julia_zakharova2 Links fixed 😄 Appreciate the find!
Happy to answer any questions or walkthrough anything you're curious about. 👍🏻