The objections a serious buyer actually raises.

What is Clyreon?

A permanent record of why your company made the decisions it made: the question, the facts it rested on, the reasoning, how sure people were, what would have proved them wrong, who disagreed, what was chosen, and what actually happened.

It is a memory of your decisions, not a tool for making them faster.

Who is it for?

Teams whose decisions take a long time to prove right or wrong. If you decide in a week and find out in a year, nobody remembers the reasoning by the time the answer arrives. Renewals are the clearest example, which is why Customer Success is usually the first team on it.

The same thing works for sales pursuits, product bets, hiring, procurement, supplier risk, legal review, budgets and escalations. Nine teams ship today, already set up rather than waiting to be configured.

Is this a CRM?

No. Clyreon sits alongside your CRM, ticketing system and analytics rather than replacing any of them. It reads facts from them and never writes anything back.

A CRM records the state of a relationship. Clyreon records the state of a belief.

Is this a knowledge base?

No. A knowledge base holds what your company knows, and it is meant to be edited as understanding improves. That is the whole point of it.

Clyreon holds what your company believed on a particular day, and it cannot be edited at all. The two have opposite requirements, which is why one system cannot honestly be both.

Is this an AI assistant?

No. Everything important works with no AI involved. Where AI is used it is optional, it works only from facts your company already recorded, and it uses a model you supply and pay for.

A model can never record a claim, say how sure you are, speak for somebody who disagreed, or decide anything. Those are not settings.

Why not just use the CRM, the docs, the chat and the dashboards we already have?

Because between them they keep the decision and almost none of the thinking. The CRM field is overwritten. The document gets edited to read better in hindsight. The chat thread is unreadable after a week. The dashboard shows the number, not what anyone concluded from it.

None of those systems is bad. None was built to hold a belief that has to survive being proved wrong. The product page sets out each case.

What exactly gets recorded?

A decision opens with the question, what is at stake, and the date it must be answered by. Against it: the facts, each a plain sentence with a source and a date; a position from a named person, with their reasoning and how sure they were, and optionally what would prove them wrong; any disagreement, which is recorded automatically when two people take different positions; the moment the decision was made; and the outcome, once it is known.

Reads, exports, refusals and admin actions are recorded too, in their own separate log.

What happens when a fact turns out to be false?

It is marked as false. The original is never deleted or rewritten, so the record of what was once believed true stays intact.

Clyreon then answers one specific question: which decisions are now worth another look. That includes decisions in other teams that never used the fact and only relied on a decision that did. Nothing is declared wrong. Whether a judgment still holds is a question only a person with the authority to judge can answer.

A fact simply replaced by a better measurement is treated differently and deliberately, because judgments made from the best available evidence at the time are not undermined by a later, better reading.

Can history be rewritten?

No, and that is the point of the product rather than a missing feature. Nothing is overwritten and nothing is deleted. Changing your mind records a second judgment beside the first, so what you originally thought, and the day you stopped thinking it, is still there.

This applies to everyone, including administrators, and including corrections made for good reasons. A system whose whole value is a trustworthy history cannot let that history be quietly tidied up.

Who can see our decisions?

Only people in your company. That boundary is enforced twice: once in the application, and once in the database itself, so a mistake in the application code still cannot leak anything. Within your company, every permission is granted explicitly rather than inherited, and reads are logged as well as changes.

Does Clyreon replace our existing systems?

No, and it is not designed to. Information flows one way: in. Clyreon never writes back to a connected system, and it never becomes the place a renewal is processed or a ticket is worked.

What if a connector we need does not exist?

Facts can always be typed in, in exactly the shape a connector would have produced, or brought in as a file. Nothing in the product needs a connector to be useful, and a connector that is not built says so rather than appearing to work.

Can we bring the decisions we have already made?

Yes. A decision log in a wiki, a spreadsheet or a ticketing system comes in as history. Each decision keeps the date it was really made, the facts cited with it, who the source says decided, and how it turned out. Imported decisions stay marked as having come from elsewhere for good, so they never read as something decided in Clyreon.

The file is checked before anything is written, and every problem is listed at once rather than one per attempt. Nothing is filled in on your behalf: a decision with no stated confidence is refused rather than guessed at, and a person this company does not recognise stays visibly unrecognised rather than being quietly attributed to whoever ran the import.

Can different teams use it?

Yes, and they share one memory rather than getting a tool each. A team in Clyreon is a name, a kind of decision and some wording. It stores nothing of its own.

That is what lets one fact support decisions in several teams, and a mistaken assumption in one function be traced to a commitment made in another.

How does security work?

Every action needs a permission that was explicitly granted. There is no implicit administrator who bypasses the check. Your data is kept apart from every other company's twice over, in the application and again in the database.

Connector credentials are encrypted, outgoing webhooks are signed so you can verify they came from us, and reads, exports and refusals are recorded in their own permanent log. Independent penetration testing and a completed SOC 2 report are scheduled before general availability and are not done today. The trust page has the detail.

How does deployment work?

You sign up and get your own workspace. There is no new deployment to stand up and no step on our side. Single sign-on and automatic user provisioning are both supported.

Running Clyreon on your own separate infrastructure is available where your compliance position requires it.

How does AI work here?

You bring your own model and your own account, and you pay your model provider directly. Clyreon bundles no model and takes no cut.

A model can draft a position for a person to review. It can never record a claim, set a confidence, speak for somebody else, or decide anything. Every claim it produces has to point at material it was actually given; anything else is thrown away before it reaches your record.

What is the design-partner programme?

Clyreon is not generally available yet. A design partner gets the full product as it exists today, direct access to the person building it rather than a support queue, and real influence over what gets built next. In return we ask for real use on real decisions and honest feedback when something is confusing or wrong.

What counts as success is agreed and written down before you start, and reviewed together halfway through and at the end. Either side can say the criteria themselves were wrong. That is a legitimate outcome, not a failure.

What does it cost?

There is no published price list, deliberately. Charging per seat rewards headcount, which is not what this product delivers, and a better answer has not been settled yet. Terms are agreed in a direct conversation and stated plainly before you onboard.

You are a small, early company. Why take the risk?

That is true and worth weighing. What is real regardless of team size: the product is built and running, your data is separated at the database level rather than only in code, and during a partnership you are talking to the person building it rather than a queue.

What is not yet true is on the trust page rather than left for you to discover.

Can we configure our own scoring or weighting?

No. Clyreon produces no score at all, because that is the thing it exists to replace. What you get instead is a record you can read directly, rather than a formula that becomes indefensible the moment nobody remembers why the weights were set that way.

What if we want AI to decide automatically?

Clyreon will not build that. A model can draft a position for a person to review. It never records a claim of its own. This is a permanent boundary, not a current limitation waiting to be lifted.

We are not ready to commit to a full rollout.

Neither are we asking for one. A partnership starts with one real decision your team is already making.

05

Something not answered here

A real address, answered by the person who built it.

If the question that matters to you is not on this page, it is probably a good question. Ask it directly rather than working around it.