What this actually is
Neuraphic is a laboratory, not a product company that also does research. The distinction decides what your week looks like. There is no roadmap to execute against and no launch date to protect. There is a short list of open problems, and the work is to close one of them well enough that somebody outside can check it.
What has come out so far is infrastructure rather than a product: an identity and account system written rather than rented, a database whose leader was killed mid-write on purpose to see what survived, a public site that contacts no host it does not run. All of it is checkable from outside, which is the standard everything here is held to. Something becomes a product when a result is solid enough to be useful to somebody who does not work here, and that is the only route to a product we have.
We are not a cybersecurity company, though this site said so until recently and you may have arrived believing it. The work touches security because a system you cannot trust under attack is a system whose limits you have not found, and the limits are the subject.
The research agenda
Three problems are on the list, and they are on it because the literature and the products in the market both stop short of them, and because there is a way for somebody outside to tell whether that has changed.
Telling a person from an automated client, where every defence in production today either lets competent automation through or makes real people prove themselves over and over, and the second failure is invisible because the people who give up are not counted. Proving a defence did not damage the thing it defends, because a filter that blocks an attack and also blocks legitimate work has moved the cost rather than removed it. Making a fix verifiable before it ships, so that the evidence travels with the patch and can be re-run by whoever receives it.
Each one carries a written condition for being solved, which means the list can be wrong in public. That is deliberate, and it is the part of this job most people find uncomfortable. If you would rather work where the goalposts are described after the result, this will not suit you.
The method goes out with the result, and what did not hold goes out in the same entry rather than in a quieter one later. There is no publishing calendar, so the gaps between entries are visible and are left visible.
The engineering work
Almost everything under the research runs on infrastructure we wrote. The authentication, the session handling, the rate limiting, the abuse prevention, the database and its failover, the deployment path: those are ours, in Go, on self-managed instances rather than managed services, with the keys held here. That is not ideology about vendors. A system you rented is a system whose failures you can only watch, and watching is not the job.
It also means the operational floor is yours. A node that dies at three in the morning dies for someone on this team, and the postmortem is published. We have written up our own worst week in public, five days during which every alarm fired correctly and none of them could reach a human, because a writeup that omits the part that makes you look bad teaches the wrong lesson with authority.
Very little is settled. There are strong opinions here but no calcified patterns, and if you join now you will set architecture and process that outlast you. That is a real offer and a real risk: nothing protects you from a decision that turns out to be wrong except the discipline of being able to tell.
Who we look for
Years of experience are close to irrelevant here and we do not filter on them. What matters is whether you can tell the difference between a result and the appearance of one, because at this size nobody is downstream of you to catch it.
We value depth over breadth. We would rather hire someone who has spent three years thinking about one hard problem than someone who has surface familiarity with twenty technologies. We value writing, not because we are a writing company, but because the ability to explain complex ideas clearly is the single best predictor of clear thinking we have found. And we value disagreement. The worst thing that can happen at a company this early is consensus for the sake of harmony. We need people who will tell us when we are wrong, and who can do it with enough rigor that we actually change our minds.
Remote-first and asynchronous by default. No mandatory meetings, no time-zone gatekeeping, and communication that happens mostly in written documents and code. This works well for people who are self-directed and badly for people who need external structure, and we would rather say that now than find out together.
One thing to know before you apply. Only one person here is named publicly: the founder. Nobody else appears on this site, in the press, or in an announcement, and we will not confirm who works here to anyone who asks. If your career depends on public attribution, this is a genuine cost and you should count it now rather than negotiate it later. What you build gets published. Your name attached to it does not, unless you ask.
Compensation reflects where we are. We are frequently not in a position to match the cash a large institution pays for a comparable seat, and we will not pretend otherwise in a first conversation. What we offer instead is equity, the scope of the work, and decisions that stay made. Neuraphic has not raised outside capital, which is why the cash number is what it is and why the equity is not diluted by anyone else. That is the deal, stated plainly.