Hello Print World

When we started building Gnaww, the first problem looked fairly obvious.

A customer says they need 500 A5 flyers, double-sided, full colour, on something that feels decent. Software should be able to take that ordinary request, work out what the customer means, spot what is still missing, and turn it into something a production system can actually use.

It can.

Then we hit the other half of the problem, which turned out to be the harder one. How does that same software understand what the print businesses on the other side can actually produce?

One has a modern API. One has a webstore. One has a spreadsheet. One has an MIS with its own terminology. And plenty have a perfectly capable factory, a website that tells you roughly what they do, and decades of production knowledge sitting inside the people who work there. They might all be able to produce the job. To software, they can look like completely different worlds, and most of them look like nothing at all.

That is the problem Gnaww is built to deal with. Not to make print tidy, but to make it legible to technology without asking the whole industry to rebuild itself first.

Gnaww is the intelligence layer for print

Gnaww is Jawwws' print intelligence and interoperability product, powered by SpecMatch™. Its job is to make print understandable to technology.

On one side, it takes incomplete, inconsistent, natural-language or system-generated demand and works out what it actually means. It preserves what is already known, identifies what is missing, asks for clarification where it needs to, and turns the result into controlled print demand.

On the other side, it works to understand production capability, wherever that capability happens to live. A catalogue, a website, a spreadsheet, an API, an MIS, a workflow system, equipment information, or the structured knowledge of people who run the floor. Gnaww normalises the meaning so software can understand what that business can genuinely do.

Then SpecMatch™ compares the two.

Gnaww understands print demand and production capability. SpecMatch compares the two.

It sounds straightforward in one sentence, but print has rarely been straightforward.

Print businesses shouldn't have to become software companies

The world is changing, there is a lot happening around AI agents, machine-readable commerce and systems that can increasingly discover and buy products without a person navigating every page. That is potentially very useful for print. It also creates an obvious problem.

If becoming visible to those systems means every print business first has to build an API, publish an MCP server, replace its MIS, rebuild its ecommerce platform and translate everything it knows into somebody else's new vocabulary, a large part of the industry starts at a disadvantage. We don't think that should be the price of entry.

A capable print business, whether it sells through a storefront or manufactures the work itself, shouldn't need to become a software company before software can understand it. Gnaww can start with the information and systems that business already has. Call it "auto on-boarding", you bring the website, catalogue, spreadsheet, API, MIS, workflow or production knowledge you already have, and Gnaww teaches systems what it means. That is why we have started describing Gnaww as the AI gateway for print businesses.

The point isn't to tell a printer to throw away 30 years of systems and adopt the Gnaww way of describing everything. The point is to understand the production meaning that already exists, keep track of where it came from, and make the useful parts available to authorised systems in a form they can work with.

You don't all need to speak the same language

Print has no shortage of terminology. Standards, supplier vocabularies, MIS fields, ecommerce product names, production terms, and the language people actually use on the factory floor. Some of it is formal, some is specific to one system, and some has grown through years of practical use because everyone involved knows exactly what it means.

Forcing all of that into one universal vocabulary is not the only way to make systems interoperable. Gnaww's approach is different.

You don't all need to speak the same language. Gnaww understands the differences.

If an external vocabulary becomes widely adopted, we can map it. If PromoStandards is useful for part of the industry, we can map it. If an MIS has its own terminology, we can map that too. And if an experienced printer explains that they can run a particular job on a particular press, within a particular sheet size and finishing route, that meaning matters just as much. The important part is understanding what those things mean, where the knowledge came from, and whether two apparently similar descriptions actually represent the same production capability. Names alone don't solve that.

Knowing what a machine is called isn't enough

This is where print gets hard for software very quickly. A substrate might run on one device at a given thickness, but only within a certain size envelope. A decoration process might work with one material but not another. A finishing operation might need to happen after another operation. A fold can change the imposed geometry, and quantity can change the sensible manufacturing route.

A machine may technically perform an operation while the site it sits in cannot complete the whole job. A substitute might be perfectly acceptable within the customer's tolerance and still need approval. One capability claim might have arrived from a supplier API this morning, another might be reviewed knowledge from the person who runs the factory.

That is a long way from teaching an AI that A4 is a paper size and CMYK is a colour model. The difficult part is relationships, dependencies, tolerances, evidence, materials, geometry, resources, sequences and constraints. That is the intelligence we are interested in.

Capability, price and turnaround, and knowing which is which

Gnaww does carry price and turnaround. It can ingest, normalise and compare them from a producer's catalogue, pricing rules, quote service or API, and SpecMatch can use them when it weighs up options. What matters is that Gnaww is honest about where each figure came from and how firm it is.

A catalogue rate is not a live quote. A calculated estimate is not a confirmed price. And being technically capable of producing something is not the same as current capacity, a delivery promise, producer acceptance, or an order. Those facts belong to the systems that actually own them, so Gnaww connects to those systems rather than inventing an answer, and every figure arrives labelled for what it is: catalogue, calculated, estimated, quote-only or confirmed. Gnaww never dresses up an estimate as a promise.

This gets more important once AI is involved, not less. A plausible answer isn't good enough when someone is about to spend money or manufacture something physical. Knowing when you don't know something is part of what makes the system useful.

AI visible = Yes

There is a small idea in the Gnaww identity that we love.

AI visible = Yes

It doesn't mean putting every printer and every piece of production information onto the public internet. Machine visibility and public visibility are different things. A business might choose to expose capability to one customer application, an internal buying agent, a private production network, a partner platform or a single integration. The important part is that an authorised system can understand it.

That is a real opportunity for the print businesses software has always found hard to see. A brilliant printer with no public API is still brilliant. A specialist producer shouldn't disappear simply because its capability lives in the wrong database. A storefront shouldn't have to join one marketplace or production network just to make what it can offer machine-readable. Gnaww lets a business be understood by technology on its own terms, without giving up its systems, its knowledge or its independence to get there.

Gnaww is already becoming a real working layer

This isn't just a diagram we have drawn for the future. The Gnaww API and developer documentation are live, and we are already working with ordinary-language and structured print requirements, controlled specifications, production capability and SpecMatch. It is also already working inside real products, with Gnaww already a plugin in Atomyx, integrated into Bytes, managing artwork with Antigro Designer, and many more.

We have also been careful to test the boundaries rather than pretend the whole print transaction belongs inside Gnaww. Editing can belong somewhere else. Preflight can belong somewhere else. Workflow and commerce can belong somewhere else. Production certainly belongs with the people and systems that actually make things. Gnaww's job is to make the meaning travelling between those systems better.

That matters because we don't think the future of print software needs another enormous platform trying to own everything. It needs systems that are much better at understanding each other.