How it works
Begin with a question about a place and something you might do. Your AI assistant prepares the simulation and helps you explore the results. You see the work in your browser, choose what to explore next and decide what to do with it.
Rosa chooses a spot for her pupusa truck
Imagine Rosa choosing where to park her pupusa truck. She has three permitted corners in mind and years of experience selling her food at a market. Which spot could work for the customers she wants to serve, at the hours she would be there?
The people, proposal and outcomes in this illustration are imagined.
01Three spots
Where should Rosa park her pupusa truck?
She has sold pupusas at a weekend market for years. Now she has a truck and three permitted spots to choose from.
Start with a question of your own.
Three spots. One food truck. Different possibilities.
Rosa is an imagined food-truck owner. This animation illustrates questions Whathappensify could help her explore.
About the illustration
The people and their movements are drawn for this explanation. The truck parks in three permitted spots, and the outdoor activity changes with the time and the work-from-home assumption. It illustrates how a question can develop; it is not a sales forecast or a recorded platform result.
Define the question before preparing the simulation
The assistant helps Rosa shape the comparison: who passes each corner, when they are there, and where else they might buy lunch. Morning commuters, office workers at lunch and weekend shoppers create different possibilities.
Rosa brings knowledge a map cannot supply. She knows when her regular customers buy lunch, which days bring people into the neighborhood and where they already like to eat. Her own records and observations help the assistant choose useful evidence and identify assumptions that need her attention.
They compare the three corners at hours that matter to her business. What happens on a day when more people work from home? The lunchtime crowd near the offices could change, and another spot might make more sense that day.
Approve the conditions and measures
In the browser, you can see the proposed setup before approving it: the area, the changes, the assumptions and the comparisons the assistant will make. You can correct it or ask for another approach.
Rosa reviews the days, the hours and the assumptions about who would be nearby. She can correct something that does not fit her experience. A change to the approved setup comes back for review; the assistant works within the scope she authorizes.
A proposal can cover one simulation or a bounded experiment with several linked runs. You see the alternatives, what each run can use, how later runs depend on earlier results and the conditions that stop the work. You approve that scope once and can watch each run as it happens. A material change outside it returns for your review.
Run the comparison and inspect its basis
The simulation makes the alternatives visible. Rosa can watch how people pass the corners at different hours and compare a busy office day with one when more people work from home. She can move the truck to another permitted spot in the comparison. The result helps her explore a choice before committing her time and ingredients.
The Epistemic Protocol Machine, or EPM, keeps the analysis connected to its evidence. Rosa can ask why one spot looks promising, where the estimate of nearby customers came from and what changes when fewer people go to the office. An answer can lead to another comparison or an observation she wants to gather.
The saved reasoning distinguishes what Rosa observed, what the simulation calculated and what the assistant concluded. She can examine the explanation and take the question further.
Follow the consequences
Rosa's choice affects more than foot traffic at one corner. The people who arrive create a queue; service moves people through it; stock and staff limits affect how quickly it clears. Her assistant connects those processes with explicit units, time windows and assumptions. She can open the model in the browser, inspect the flows and equations, change a rule and prepare another comparison.
The same approach connects long-term change with events that unfold in minutes. A town can compare how trees grow and shade a waiting area through the years, or how a transport investment changes trips and how those trips affect the next round of demand. Each part keeps its own clock, inputs and result, and the connections show what passed between them.
Ask what the choice depends on
A busy day and a quiet day need not lead to the same choice. The assistant varies declared assumptions, repeats processes that involve chance and shows the spread of outcomes. It keeps uncertainty about the inputs, randomness, numerical calculation and model structure visible separately.
Rosa can ask which assumptions drive the difference, whether a different service rule performs better, or whether counting customers at a particular hour could change her decision. The platform compares the stated options under the same conditions and keeps unsuccessful runs in the account. She can see both the overall result and effects on particular groups.
Bring different perspectives to the same result
The busiest spot is not automatically the one Rosa wants. She may care about serving regular neighbors, keeping prices within reach or finishing in time to collect her child. A customer using a wheelchair needs a place to approach the truck and order. A neighbor may care about customers blocking the sidewalk.
Perspectives make those points of view explicit: what matters, why it matters and how the simulation represents it. Rosa can examine the alternatives through her priorities and other people's concerns. Reading a result through a Perspective changes how it is read, not the result itself. A concern the simulation cannot represent remains visible and can shape the next question.
This gives her a clearer basis for deciding what to explore next. If she needs to understand an effect the comparison did not examine, her assistant can propose the additional work it requires.
Decide what to share and what to do next
Rosa can keep her business plans private, return to the comparison or explore another arrangement. If she chooses to share suitable work, other people can examine its basis and develop alternatives, subject to the permissions on its inputs.
Another food-truck owner might make a fork—a version they can change—and attach a Perspective about the customers they want to serve. Someone in a different neighborhood could emulate Rosa's approach using their own observations. Sharing a method need not mean publishing her sales records.
Later, Rosa can bring back what she learns and compare it with what she expected. One question becomes a foundation for the next. The same ability can help people develop a service together, explore a change to their neighborhood or simulate back when a public proposal affects their lives.
Explore more possibilities · What it can simulate · How your assistant connects