What we promise, and what we will not do.
Most companies publish values. This is narrower and more useful: the specific commitments we make about how we build with AI, how we treat you when you need help, and what we are personally on the hook for. Every one of them is something you can hold us to, and every one applies to every product we ship.
AI is a great tool. We develop and use it responsibly.
We are not hesitant about AI and we are not evangelists for it either. It is the most capable tool we have had in a long time, it is in our products and in how we build them, and the whole question is whether the people using it are honest about what it is doing. Here is how we hold that line.
- A tool, not an authority
- AI is one of the best tools our industry has been handed, and we use it every day — to build faster, to catch what we would have missed, and to give small teams reach they could not otherwise afford. It does not get the last word. A person reviews what ships, and a person is accountable for it.
- Responsibly, in the boring sense
- Responsible use is not a statement of values, it is a working habit: check the output, understand the thing before shipping it, and do not put a model in a position where nobody would notice it was wrong. That is the standard we hold ourselves to on our own products and on client work.
- You always know when it is AI
- Where a surface is AI-driven, we say so plainly in the interface. No agent presented as a person, no generated work handed over as if a human checked it line by line when nobody did.
- Your work is not raw material
- What you put into our products is yours. We do not treat customer content as a dataset to be mined, sold, or fed into training something else, and we will not start without asking you first, in plain language, with a real choice attached.
- It asks before it acts
- Anything that can change your data, spend your money, or speak in your name should propose rather than assume. We would rather ship an assistant that pauses one time too many than one that is confidently irreversible.
- We say what it cannot do
- We do not market capability we have not shipped, and when a model is the wrong instrument for a job we will tell you so instead of selling you the version with AI in the name. The same honesty we apply to our security gaps applies here.
- Novelty is not a reason
- A feature being possible is not an argument for building it. If it does not leave the person using it more capable — faster, clearer, with their time back — it does not ship, however good the demo looks.
Support you would actually recommend.
Support is where software companies reveal what they think of the people paying them. Ours is deliberately unfashionable: a small number of humans who answer quickly, know the product, and are allowed to fix things.
- A person answers
- Every support request reaches a human being who can actually change something. No deflection bot at the front door, no ticket maze designed to make you give up, no tier one that exists to read you the documentation you already read.
- Inside one business day
- We answer in writing, within one business day, signed by name. If the fix is going to take longer than that, the first reply tells you what is happening and when to expect the next one — silence is the part that actually costs you.
- The people who built it
- We are a small company and that is the advantage: the person answering you has read the code, or wrote it. Questions get answered from the source rather than from a script.
- Help is not a paid tier
- Getting your account working, understanding a feature, recovering from something we broke — that is support, not an upsell. We do not put the answer to "why is this not working" behind a plan upgrade.
- No question is too small
- If something is confusing enough to ask about, it is a problem with our product, not with you. We would rather answer it four times and then fix the thing that made it necessary.
- We will tell you when it is no
- If we cannot do what you need, or you would be better served elsewhere, we will say so directly and point you somewhere better. A customer we kept by being vague is not a customer we wanted.
- When we break it, we say so
- Outages, regressions and mistakes get told to you by us, in plain terms, before you have to come asking. We write them up afterwards on the changelog rather than letting them quietly disappear.
The personal part.
I started Stacklumen because the gap between what enterprise software can do and what a small business can actually get its hands on is indefensible. The capability exists. It is just priced, packaged and gatekept away from the people who would do the most with it. Closing that gap is the whole job.
So the promise is simple. We only ship things we would be glad to hand to someone we care about. If it is not good enough for a friend's business, it is not good enough to sell. That rule has killed more features here than any roadmap meeting ever has.
Software should leave the person using it better off — more capable at the end of the day than they were at the start, with fewer things to worry about and more of their time back. That is what "makes the world better" means at our scale. Not a slogan about changing the world: a small business that is steadier, an owner who sleeps, work that comes back to them because their site finally does its job.
We will price honestly, we will not lock you in, and your data leaves with you if you go. When we get something wrong — and we will — we will say so first, fix it fast, and tell you what we changed so it does not happen twice.
If we ever fall short of any of this, write to me directly. Not a form, not a queue. Me.
The same commitments, on every product.
We build more than one thing, and a commitment that only holds on the flagship is marketing. This page exists on each product, written in that product's own terms and promising exactly the same things.
Stacklumen
The agency platform — client work, the Hub, and the Terminal site builder.
You are hereBaselumen
The component library, client portals, and the tool bench.
Read its commitments →Operalumen
Agentic orchestration for design and print.
Read its commitments →Holding us to it
If anything on this page does not match your experience of working with us, that is a bug and we want to hear about it: support@stacklumen.com, or iain@stacklumen.com if you would rather it went straight to the top. The enforceable version of our obligations lives in our legal documents, and how we handle your data is set out on security.