SysSimulator is a browser-native system design simulator built by Ronit Dahiya. It models distributed architectures as discrete-event systems — load balancers, databases, caches, message queues — and runs traffic through them at up to 100,000 RPS without any backend infrastructure. Engineers use it for system design interview preparation, architecture review, and resilience planning. 57 pre-built blueprints. 28 chaos scenarios. AWS cost estimation. No signup. No infrastructure. Runs entirely in your browser.
Most system design tools are drawing tools. They let you arrange boxes and arrows into diagrams that look like architectures. SysSimulator is different: it executes those architectures.
Drop a load balancer, three app servers, a Redis cache, and a PostgreSQL primary with two read replicas onto the canvas. Set traffic to 10,000 RPS. Watch p99 latency, error rates, and bottleneck identification update in real time. Then inject a cache stampede and narrate what cascades — because that exact narration is what separates a hire from a no-hire in a system design interview.
The simulation is powered by a Rust discrete-event engine compiled to WebAssembly. It runs at full CPU speed inside the browser tab. Your architecture never leaves your machine. There is no server receiving your design, no account to create, no subscription gating the chaos panel. The entire product is a static file deployment.
Static diagrams don't fail. That's the problem.
Every system design resource teaches you to draw boxes. Draw a load balancer. Draw a cache. Draw read replicas. But systems don't live as diagrams — they live as traffic, failure modes, and cascading latency. The skill interviewers actually probe is whether you can reason about what happens when your cache evicts 50 million keys simultaneously, or when your primary database becomes unreachable at 3am during Black Friday traffic.
I built SysSimulator because I wanted to practice that reasoning without setting up infrastructure. The goal was a canvas where you could design a system in five minutes and break it in the next five — and walk away with specific numbers to talk about: p99 at X RPS, error rate after Y chaos event, monthly cost at Z scale.
Rust and WebAssembly were the natural choice for the simulation engine. The browser tab is the deployment target — no server latency, no rate limits, no cloud costs. Rust gives the engine the performance headroom to simulate 100,000 concurrent virtual requests without dropping frames. WASM gives it cross-platform portability without a runtime dependency. Flutter Web handles the canvas and UI, which needed a widget system that could manage hundreds of draggable, selectable, configurable components without performance degradation.
The engine uses discrete-event simulation — a technique used in production capacity planning, network modeling, and queuing theory. Rather than approximating behavior with averages, it processes individual events in time order: request arrives, request enters load balancer queue, request routes to app server, app server hits cache, cache returns hit or miss, response completes. Each component has configurable throughput limits, latency distributions, and replica counts.
This produces numbers that behave like real systems. Connection pool exhaustion, queue depth buildup under sustained load, the difference between a cache hit rate of 98% and 95% — these are not illustrative animations. They are outputs of a simulation model grounded in real-world component benchmarks.
The privacy model is a direct consequence of the architecture: because the simulation runs entirely in WebAssembly inside the browser, your system design is never transmitted anywhere. There is no telemetry on the canvas. No design you draw is visible to anyone but you.
Read the full technical explanation of the simulation engine and component model: Distributed Systems Simulator →
Browse all 57 architecture blueprints covering microservices, event-driven systems, MCP agent architectures, and more: Architecture Blueprints →
Browse 28 in-depth system design guides covering CAP theorem, consistent hashing, database sharding, and full system design walkthroughs: System Design Learning Hub →
Prepare specifically for FAANG and top-tier system design interviews with narration guides and common question walkthroughs: Interview Prep Guide →
Builder
Builder of SysSimulator — Rust/WASM simulation engine, Flutter Web canvas, chaos engineering, and 57 architecture blueprints.