TL;DR
Get tech for your team delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
The AI Company Emulator is publishing a day-by-day replay of an AI team running GewerkTon, a construction-site app. The run began from GewerkTon’s real state, but all activity and results after day 0—including a simulated first paid licence on day 44—are not real business results.
The AI Company Emulator has published a replay of an AI team running GewerkTon, a construction-site app, through 44 simulated business days. The run starts from the product’s real beta-stage position, but all subsequent customers, deals, decisions and figures are simulated, so its day-44 paid licence does not represent actual GewerkTon revenue or traction.
The replay begins with GewerkTon’s real starting state: one founder, an experienced site manager testing the app in beta, and no customers. From day 1 onward, the company’s activity is emulated. The site presents each business day as a sequence of decisions and actions, with a live feed, an office map showing the simulated team at work, and a timeline for moving between days.
The simulated team has six AI employees across five roles: product, engineering, pilot success, business development and finance, with two engineering agents. In the run through day 44, it wins its first pilot on day 6, ships a requested feature on day 16 and converts a pilot into a paid licence on day 44. The replay reports 13 pilots won, 10 active pilots and 48 releases by day 44; those totals describe the simulation, not real GewerkTon operations.
The record includes setbacks as well as milestones. According to the replay, reviews blocked all proposed work at one point, while offers were not recorded. The founder then issued short directives, which appear in the decision feed. The project presents these moments as a way to observe where the simulated team stalls and what changes its behavior, making AI leadership in startup success a relevant point of comparison. It also says each business day corresponds to a git commit, making the sequence of simulated work inspectable.
A 44-Day Startup Replay · The Boundary Between Model and Market
The Daily Decisions Behind An AI Team’s Startup Journey
An observable replay follows an AI team running GewerkTon, a construction-site app. The starting point is real; every business event and result after day 0 is simulated.
“Everything after day 0 is emulated.”
Thorsten Meyer AI01 / How the replay works
Real starting state. Modeled company journey.
GewerkTon’s beta-stage position gives the run its initial conditions: one founder, an experienced site manager testing the app, and no customers. From day 1, the team’s work and the surrounding business environment are emulated.
A beta construction app
GewerkTon is described as serving construction-site managers, tradespeople and facility operators. Its real beta state anchors the simulation.
Six AI employees
Five roles cover product, engineering, pilot success, business development and finance. Engineering has two agents.
Decisions over time
A live feed, office map and day-by-day timeline show the simulated team at work. Each business day is said to correspond to a git commit.
02 / Milestones in the run
From first pilot to paid licence
The reported milestones describe events inside the emulator. They make the sequence legible, while remaining modeled outcomes rather than real GewerkTon business results.
First pilot won
The simulated team secures its first pilot relationship.
Requested feature ships
A feature requested by a pilot user is delivered in the replay.
Pilot converts
A pilot becomes a paid licence within the simulated run.
Day 44 · Simulation state
What the replay counts
These totals summarize modeled activity through day 44. They should be read as emulator state, not as customer traction, releases or revenue achieved in live operations.
03 / What it can show
A window into modeled management
The public, day-by-day format lets readers trace how a multi-role AI workforce responds to feature requests, pilot relationships and business development—rather than judging it from one answer or demonstration.
Follow cause and effect
Readers can inspect the team’s actions and consequences across a sequence of simulated business days.
See where work stalls
The replay says reviews once blocked all proposed work and offers were not recorded. The founder responded with short directives in the decision feed.
Observe behavior
The record may show how agents act within the emulator’s rules and what changes their behavior. It does not establish live-company performance.
04 / Where simulation meets reality
Useful evidence has a boundary
The replay is an observable experiment in modeled company behavior, not a customer case study. Its post-starting-point figures do not verify commercial outcomes for GewerkTon.
The initial condition is GewerkTon’s real beta-stage state: one founder, a site manager testing the app and no customers.
Pilots, prospects, deals, decisions, releases and the paid licence are modeled. Actual GewerkTon traction after the beta starting point is not specified.
No real-world comparison shows how the AI team’s decisions would perform against human-run operations or whether simulated customers would behave the same way in a live market.
The available description does not explain model selection and configuration, management-quality scoring, or how closely simulated buyers reflect real ones.
Traceability / What connects the story
Starting state → simulated work → inspectable record
Firmulate is the emulator behind the replay. Thorsten Meyer AI describes it as running AI models as complete companies—with crises, financial mechanics and incentives—and scoring management quality rather than chat quality.
GewerkTon beta
The real product state initializes the run.
Firmulate
AI roles act within a modeled business environment.
Daily replay
Feed, timeline and commits expose the simulated sequence.
05 / What comes next
Follow the next simulated days
The site says it adds replay days daily. Later entries may reveal further milestones, setbacks or founder interventions. Any new result carries the same boundary: the starting point is real, while activity and outcomes after day 0 remain simulated.
What is the AI Company Emulator replay?
A day-by-day simulation of an AI team running GewerkTon through modeled company decisions and events.
Are the pilots and paid licence real?
No. The run starts from GewerkTon’s real beta state, but pilots, deals and later results are simulated.
What milestones does it report?
A first pilot on day 6, a requested feature shipped on day 16, and a simulated paid licence on day 44.
What is Firmulate?
The emulator behind the replay, described by Thorsten Meyer AI as running AI models as companies and scoring management quality.
What the Startup Replay Can Show
The replay gives readers a view of how a multi-role AI workforce might handle a startup’s recurring work, including feature requests, pilot relationships and business development. Its public, day-by-day format makes it possible to follow decisions and consequences over time, rather than judging the system from a single answer or demonstration.
That view has limits. Because activity after day 0 is simulated, the reported pilots, releases and paid licence do not establish that AI employees can produce those results in a live company. The replay is better understood as an observable experiment in modeled company behavior: it can show how agents respond within the emulator’s rules, while leaving real-world performance and commercial outcomes unproven.
construction site management software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From GewerkTon Beta to Simulation
GewerkTon is described as an app for construction-site managers, tradespeople and facility operators. The emulator uses its actual starting state as the initial condition, then runs an AI team against a simulated business environment. Prospects may respond or remain silent; pilot users can request features with deadlines; and customers in the simulation can react when commitments slip.
The underlying emulator is called Firmulate. Thorsten Meyer AI describes it as a system that runs AI models as complete companies, with crises, financial mechanics and incentives, and says it scores management quality rather than chat quality. The GewerkTon replay is explicitly not a customer case study: its post-starting-point figures are modeled outcomes.
“Everything after day 0 is emulated.”
— Thorsten Meyer AI
As an affiliate, we earn on qualifying purchases.
Where Simulation Meets Reality
The replay does not establish whether its simulated pilots, feature requests, customer reactions or licence conversion would occur in a live market. It also does not provide a real-world comparison showing how the AI team’s decisions would perform against human-run operations. The figures reported for day 44 are simulation state, and actual GewerkTon customer traction after the beta starting point is not specified.
The available description also does not detail how Firmulate selects or configures the models, how its management-quality scoring works, or how closely its simulated prospects and customers reflect real buyers. Those details would help readers assess how far the replay’s behavior can be generalized beyond this particular run.
construction app for site managers
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Replay’s Next Simulated Days
The site says it adds new replay days daily, so readers can follow the emulated team beyond day 44 as it encounters further decisions and simulated business conditions. The next development is whether later entries show new milestones, setbacks or founder interventions.
Any subsequent results should be read with the same boundary in mind: the run starts from GewerkTon’s real beta state, while activity and outcomes after day 0 remain simulated. The replay’s ongoing record may offer more evidence about how the emulator behaves, but it does not by itself confirm commercial results for GewerkTon.
Source: Thorsten Meyer AI
software development tracking tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the AI Company Emulator replay?
It is a day-by-day simulation of an AI team running GewerkTon, a construction-site app, through modeled company decisions and events.
Are the pilots and paid licence real?
No. The replay starts from GewerkTon’s real beta-stage state, but pilots, deals and other results after day 0 are simulated.
What milestones does the replay report?
In the run through day 44, the simulated team wins a first pilot on day 6, ships a requested feature on day 16 and converts a pilot into a paid licence on day 44.
What is Firmulate?
Firmulate is the emulator behind the replay. Thorsten Meyer AI describes it as a system for running AI models as companies and scoring management quality.
Source: Thorsten Meyer AI
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
