Dana Rivera
“I have rebuilt this system twice already. Nobody has asked what we learned the last time.”
Leadership simulation · free in the browser
Lead Through Noise hands you a simulated team where every person has a personality, skills, incentives and a life outside work, and 25 initiatives that go sideways the way real ones do. Announce the reorg. Push the deadline. Watch who checks out, who steps up and who quits, then trace exactly why, week by week. Your mistakes stay in the sim.
NO ACCOUNT · NO INSTALL · ONE RUN ≈ 15 MINUTES · BEST ON DESKTOP
How a run plays
A full run takes ten to twenty minutes. You make the calls a real leader makes, then the engine shows you how they land on twelve very different people.
Generate 7 to 35 people across four departments. Click anyone to read their personality, skills and what they actually care about.
Choose from 25 scenarios drawn from real business life. Migrate the architecture. Cut 15% of headcount. Survive a product pivot.
Ten levers, each a real tension. Urgency or patience. Announce or build consensus. Start from an archetype or tune your own.
The simulation streams week by week. People adapt, resist, hit personal emergencies and sometimes quit. Adjust levers mid-run.
The game names the person you lost and the week you lost them, then asks what you would do differently. Then you replay.
The noise has names
Team members are generated with OCEAN personality traits, correlated skills and personal incentives. Each week the most interesting people get a one-line inner monologue, written by a small language model running locally with the simulation.
“I have rebuilt this system twice already. Nobody has asked what we learned the last time.”
“Give me the goal and get out of the way. This is the most room I have had in years.”
“My dad is back in the hospital. I hear the deadline. I just cannot hold both this week.”
These three are sample output from a real run. Your team will be different every game, and life events like Amara’s arrive whether or not your timeline has room for them.
Traceable, not vibes
No hidden dice rolls and no black-box AI deciding who quits. The engine combines three published models of how opinions move through groups, and the debrief lets you click any person and follow the numbers.
Dana Rivera departed week 14. Her alignment fell from 62% to 34% in the three weeks after you raised urgency without touching transparency. What would you have done differently?
Your controls
There is no correct setting. Urgency ships faster and burns people out. Patience protects wellbeing and lets a deadline slip. The game exists so you can feel those tradeoffs before your real team does.
Fair questions
Yes. The game runs in your browser with no account, no install and no payment. The signup form on this page is only for launch updates and is optional.
No. All outcomes come from the opinion-dynamics engine, which is deterministic for a given seed and covered by 59 verification tests. A small local language model (Qwen3-0.6B via llama.cpp) only writes the one-line inner monologues, and if it is slow or unavailable the game falls back to templates without changing any outcome.
Not well yet. The game is a dense three-panel control room and currently needs a desktop-sized screen. A phone-friendly layout is on the list; the signup form below is the best way to hear when it lands.
Each team is procedurally generated: OCEAN personality traits, correlated skills, incentives and reporting lines across four departments. The same seed always produces the same team, so a debrief can always be replayed and inspected.
Branching timelines, so you can fork a run at any week and compare two approaches side by side, plus decision prompts mid-run and a persistent leader profile. None of that has shipped yet, which is exactly what the launch list below is for.
Stay on the list
Play today without signing up for anything. Leave an email if you want to know when branching timelines, mobile play and the leader profile ship.