In January we closed an angel round. We raised $1.5 million. This is the first time we have announced it publicly, so this post is partly that announcement and mostly an explanation of what the money is for and why we raised the amount we did rather than going for more.
Why We Raised What We Raised
There is a version of this studio that raises a much larger round right now and builds faster. We chose not to do that, and the choice was deliberate.
The game we are building requires getting one specific thing right before it is worth anything. The adaptive AI system needs to actually work: it needs to produce opponents that feel genuinely different from scripted AI in a way that competitive players recognize and care about. The matchmaking system needs to keep lobbies fair in a way that does not feel mechanical. If those two systems work, we have something. If they do not, we have nothing, regardless of how much we raised.
Getting those systems right requires iteration time, not headcount. We have four people who understand the problem. Adding eight more people who understand it less would not get us to the right answer faster. It would get us to a faster answer that might be wrong, and wrong answers in game development are expensive to undo.
$1.5 million gives us roughly 18 months of runway at our current rate to continue the work we have been doing since 2023: building the system, running playtest sessions, iterating based on what we observe, and shipping when it is ready. That is the constraint we are optimizing under.
What the Money Is Specifically For
About half of the funding goes to operating costs we were already carrying: the Rivington Street studio space, compute for the server infrastructure we run playtests on, and the hardware we use for development. These are costs that exist regardless of what we build and were previously covered by personal runway.
The second half funds three things that were previously constrained by budget. The first is the expanded playtest program. Closed playtesting with recruited participants is not free. The sessions we have been running since late 2025 have been limited in scale because of what they cost. We now have the budget to run them at the scale they need to be run at to generate meaningful calibration data across the full player skill range.
The second is the inference infrastructure. Running a live ML model on game server hardware at scale requires compute that we were simulating with approximations before. The current architecture we have documented in other posts is architecturally correct, but we have been running it on a small server footprint with a limited number of concurrent sessions. Scaling that to the player count we are targeting at launch requires infrastructure investment that the angel round makes possible.
The third is one additional hire. We need someone with network engineering depth to harden the backend for the kind of load that comes with even a modest early-access launch. The four of us can build the game. We need one more person to make sure the server infrastructure does not fall over when real players start using it.
What This Is Not For
We are not using this round to accelerate launch by cutting scope. The temptation with new funding is to interpret it as permission to move faster by doing less. We are specifically not doing that. The feature set of the game at launch is what we planned. The adaptive AI needs to work. The matchmaking needs to work. The map set needs to be complete. We are not trading any of those against timeline.
We are also not using this to change what the game is. The game we pitched to angels is the same game we have been building since the Rivington Street space opened. A competitive multiplayer shooter where the AI adapts to how you play and matchmaking reads skill in real time. That has not changed and will not change because we now have more money in the bank.
What Comes Next
The expanded playtest program is already running. We recruited the first cohort of the wider program in late January, and the sessions are generating the data we expected them to generate. The infrastructure work is underway. We expect to have the server footprint in the state we need it to be in by late Q2.
The early access waitlist is live. If you are a competitive FPS player who wants to be in the first group of real players through the door, the sign-up is on this site. We are not going to be precious about what counts as ready. We will open early access when the core systems are working at the quality level we set out to hit, and not before.
We are grateful to the people who backed us. They backed a small studio with a very specific claim about what competitive gaming should feel like, and they did it based on what we showed them in our early playtest sessions. We intend to be worth that bet.