The technology behind the Grid.

01  The problem

Anything that introduces people has the same weakness. It decides that two or more parties fit, tells them so, and then stops paying attention. From that moment the recommendation is a statement about the past.

With two parties that is survivable, because the two of them find out for themselves. With four it is not. An arrangement that needs a company, an operator, a fund and a route to market only works while all four are still what the arrangement assumed. If one of them changes and nobody notices, three people spend a month on something that stopped being available in week one.

The obvious fix makes it worse. Telling everybody why an opportunity disappeared means publishing a fact about a person or a company that they did not agree to publish: that a fund has stopped deploying, that an operator has taken a job, that a business has lost the contract the plan depended on. So platforms tend to choose between being wrong and being indiscreet.

02  The invention

GB2619192.4 is a network system for managing multi-party arrangements. In plain terms it treats an opportunity as something that stays true or stops being true, rather than as a message that was sent once.

An arrangement is assembled from parties who each qualify for a specific part in it. It is valid only while every one of those qualifications still holds. When one stops holding, the arrangement is withdrawn before anybody acts on it, and it is withdrawn without disclosing to the other parties whose circumstances changed.

That is the whole of what is described publicly, and it is deliberate. How qualification is expressed, evaluated and re-evaluated is the subject of the application and is not published here.

03  Multi-party arrangements

The unit is the arrangement, not the introduction. Four parties in one arrangement is a different object from six pairwise introductions, and it behaves differently:

• It has named parts, and each party is in it for a specific part rather than in general.
• It is valid or invalid as a whole. Three qualifying parties out of four is not three quarters of an opportunity.
• It has a first step, so it can be acted on rather than considered.
• It has a lifetime, and the end of that lifetime is an event the system produces rather than something the parties discover.

04  Continuous qualification

Qualification is a state, not a check that happened at the moment of matching. A party who qualified for a part on Monday may not qualify for it on Thursday, and the system is built on the assumption that this is normal rather than exceptional.

The consequence that matters to a member is simple. Nothing GridHQ shows you is historic. If an opportunity is on your screen it is because it was viable when the screen was drawn, and if it stops being viable it leaves.

05  Withdrawal without exposure

When an arrangement is withdrawn the other parties learn one thing: that it is no longer available. They do not learn which party changed, and they do not learn what changed about them.

This is a requirement rather than a courtesy. An ecosystem only works if participating in it is safe, and it is not safe if the price of being in an opportunity is that your circumstances become visible to everybody else in it the moment they move. A fund that pauses deployment, an operator who takes a role and a company that loses a contract all have a reasonable expectation that the platform does not announce it on their behalf.

06  What GridHQ uses it for

Inside the platform this is the mechanism behind the opportunity engine described on the front page. It is what makes the difference between a feed of suggestions and a single move that is actually available today.

The same property applies wherever an outcome depends on several parties at once, which in a venture ecosystem is most of the interesting cases: a funding round with a lead and a follower, a hire that depends on a raise closing, a route to market that only works while a commercial agreement stands, a grant that requires a consortium.

07  Ecosystem infrastructure

Continuous qualification depends on knowing that a party is who they claim to be, and GridHQ does not solve that itself. It runs alongside infrastructure the group already operates:

• Confrmo — identity. Establishes that a person or a company is who they claim to be, and keeps checking that it is still true.
• VCode — verification. Visual verification infrastructure. A mark on a document or an object that anyone can check with a phone.
• Couri — communications. Correspondence that carries proof of who sent it, so an introduction cannot be impersonated after the fact.

Those systems carry their own patents and their own applications, held by their own companies. GB2619192.4 covers what GridHQ does, and nothing in it extends to them. Equally, a filing held by Confrmo or VCode is not a filing held by GridHQ, and this site does not present them as one estate.