This is not a sales page; it is a disqualification list. Starting with an organisation that does not fit burns both sides' time, so we would rather make it easy for you to rule yourself out. Below: first the conditions where it fits, then the conditions where it does not. The second list is as real as the first.
If most of the following describe you, 361 may be the right tool. Not all of them need to be true — but if fewer than half are, it is not worth pushing.
361 aims to attach as a separate layer on top of your existing business system rather than take its place. If you are not looking for a rip-and-replace project, you are in the right place.
Where a regulatory expectation, contractual duty or internal policy prevents data from leaving, local AI running inside the machine and the localOnly switch address that need directly.
When a process lives in three systems and the gaps are filled by email, spreadsheets and phone calls, the gain is right there. The value of 361 is not in speeding up one task but in closing the gap between them.
If you must be able to answer "what did the AI see, what did it do, who approved it": role-based access and row-level security apply to AI too, and every action is written to an immutable audit trail.
If the expectation is opening records, producing documents, routing approvals and answering through channels rather than generating text, you need something that calls tools and can be bound to human approval.
If a process particular to your organisation has no equivalent in an off-the-shelf product, defining it on a platform is faster than building a new product.
When "who should see what" is a genuine problem for you, having the authorization model inside the platform makes a serious difference.
A POC only works if someone inside the organisation owns the process and real data is used. Without that, the result will not convince you.
We are not softening this list. If one of the following is your situation, another solution is probably better — and saying so is part of our job.
Role separation, authorization policy, approval flows and audit trails are a cost; you only get the return when several people touch the same data with different permissions. In a one-person business you pay that cost and never see the benefit. A simple off-the-shelf tool is the better answer.
We do not claim to beat a mature product that does one job well. If you need to issue invoices, collect one kind of form or pull one report, use the tool that does it. Our value appears when jobs connect to each other; with no second job to connect, there is no value.
The strongest side of 361 is built on where data sits and on AI operating under authorization. If that is not a concern for you, you would be buying that strength without using it — something lighter will serve you better.
AI does not fix an undefined process; it accelerates whatever the process already does. Automating a tangled process means producing the tangle faster. Someone has to own the definition first; without that person the project becomes an installation with no owner.
The default mode is human-approved autonomy: AI proposes, a person approves. Full autonomy is enabled deliberately, per channel and per action, with an undo path defined. If you want unsupervised automation, that expectation is not met here.
The server side runs on Windows Server. That is an architectural fact rather than a preference. If your corporate standard does not allow it, nothing starts before the deployment model is settled.
Record history, a document archive, a knowledge base — without material for the AI to work on, what you get is an empty interface. Data first, AI second.
361 is a living platform: definitions change as processes change, governance is tuned as usage grows. An installation that is set up and abandoned loses its value over time. "Install it, hand it over, let's never speak again" does not fit this model.
You will not find a price figure on this site; a number given before scope is defined misleads one of us. If your evaluation is not based on scope, the comparison will not be sound in the first place.
Some situations cannot be settled in one sentence. For each of these we have written the question that decides it.
The deciding question: within a year, how many people will touch the same data with different permissions? If the answer is single digits, do not rush; if it is double digits, establishing the authorization model early is cheaper than migrating later.
The deciding question: can you allocate a server with 16 GB RAM and 8 cores? That is where local AI starts. If not, the hardware conversation comes first and the AI conversation only makes sense after it.
The deciding question: can the data be reached through an API, the database or a regular file transfer? If none of those exist there is no surface to attach to, and without an integration surface no layer can be built.
The deciding question: will the output trigger a workflow, or stay as text? If you only need text extraction, a narrower tool is enough; if the output feeds a record, an approval or a process, the platform makes sense.
The deciding question: why did it fail — model quality, access to data, or ownership? If the cause was ownership, a new platform will not fix it. The other two are worth discussing.
The deciding question: who will evaluate the POC result and say "continue"? If that person is not identified, the POC succeeds technically and goes nowhere organisationally.
You do not have to show this to anyone; answer honestly to yourself.
If not, the topic is not 361 but the underlying system first.
If not, the area of gain stays narrow.
If not, you will not be using the strongest part.
If not, a lighter tool may be enough.
If not, the POC result will not convince you.
If not: data first, AI second.
Four or more "yes" answers make a POC worth discussing. Two or fewer means this is not the moment — and telling you that is more useful than trying to convince you. If you are unsure, the maturity test produces a more detailed picture in five minutes.
Five minutes to map your organisation's readiness; the result screen tells you which step comes first.
SaaS, private cloud, on-premise and air-gap; what works in each and what does not.
Certificates, data flow, isolation, encryption, backup, subprocessors and incident notification.
Your data, your environment. If the result does not convince you, there is no obligation to continue.
What was done at organisations of similar size facing similar problems.
If you could not place yourself in the lists above, write to us; if we think it is not a fit, we will say that too.
Describe your situation and we will assess it against the list above and give you an honest answer. If we think it fits, we will show you on your own data in a two-week POC.