Presses have never been faster or smarter, so why do so many print businesses still feel a step behind? In the latest FuturePrint podcast, Elena Knight spoke to Diana Pascual, General Manager, HP Industrial Print Software and Solutions, about where the real bottlenecks are, what AI can reliably do today, and why homegrown systems can turn from an advantage into a liability.
Diana describes herself as 'a software person at heart'. She joined HP almost 20 years ago as a student in the large format firmware team. She trained as a computer scientist and completed a master's in robotics and AI at a time when, as she puts it, the field was 'super theoretical, super experimental'. Roles in QA, colour and applications followed, bringing her closer to customers and eventually to some of HP's first factory automation projects in large format.
This is where a childhood fascination came back into focus. On a school trip to a car factory in Spain as a young teenager, she was struck by 'the robots, the power of the production line, the processes, the oversight'. Years later, she helped build some of the first print-on-demand assembly lines, where the printer was 'only one piece of the assembly line'. Her job was to design the processes and the software that kept it running and moved each job on to the next step.
She eventually moved into software full-time seven years ago, after HP acquired OneFlow Systems, the business behind HP Site Flow, and today she runs HP's industrial print software business. 'I felt like I've been preparing my whole life for this position,' she said. 'I always debated between the two sides of myself, the business and the technical. This one is really giving me both.'
She was also there at the very start of HP PrintOS, which turned ten this year. Diana remembers the first conversations well. Outside print, people were already used to apps, logins and data in the cloud. In print, 'there was no other vendor that had any cloud telemetry'. She also remembers that when the idea went to HP's board, nobody needed convincing.
Diana feels that, the platform is as much about the future as the past. 'PrintOS is turning ten, but it's going to be 20 and it's going to be 30,' she said. As print businesses connect more tools and partners, the platform becomes even more important.
“The presses are there, just waiting for volume to come.”
When we met at FESPA in Barcelona earlier this year, Diana told us the real challenge now is managing the complexity around printing.
Most PSPs and converters, she explained, effectively have two floors. One is the business floor, where orders, customers, sales and invoicing are handled. The other is the production floor. Prepress began its automation journey 20 to 25 years ago and is now highly automated, but everything before and after it is a different story. Orders are getting shorter and more frequent, and they arrive by email, through file shares, via APIs, or all of the above.
'The presses are there, just waiting for volume to come,' she said. 'Maybe the volume is at the entry, but there are stages in the middle where things get stuck, preventing printing from happening.' Someone recently told Diana, 'You are not just a software business, you enable printing,' and it has clearly struck a chord with her. The aim, she said, is to help customers 'scale profitably without adding more people'.
Most of us will recognise the risk of the one person who knows how everything works but has never written it down. Diana sees this happening in two areas. On the shop floor, PSPs are professionalising quickly, bringing in manufacturing disciplines such as documented processes and proper shift handovers.
The second area is IT. 'When we think about who our competitors are, our key competitors are the homegrown systems,' she said. Many businesses began their digitalisation journey with one IT person, or a small team, who connected the ERP, MIS and CRM and built bespoke tools around them. When those people leave, the business can be left with undocumented software at the core of the factory.
She is careful not to call homegrown systems a mistake. Bespoke operations can be a real differentiator. But there can come a point where they become a liability. Security can slip, technology moves on, and modernising 20 years of code could take '25 engineers'. 'They are not an IT company, they are a printing company,' Diana said. 'If you make it that big, you need to become an IT company to keep up with it.'
“Do it right, not fast.”
Diana sees AI as 'smart computation over data', so it works best where data is already structured. Scheduling is the clearest example, but always with a person in the loop. Customers don't want work moving from one press to another without their knowledge. 'It's a judgment issue,' she said. 'What you want is AI to simplify all your thinking and speed up your judgment and your decision-making.'
Diana is also seeing more printers plug public AI models straight into their homegrown systems, and her advice is to be cautious. Doing so can mean exposing business data, and building the guardrails to keep it private, secure and ethical is 'super, super challenging'. Her first recommendation is an off-the-shelf solution with that protection built in. If businesses do build their own: 'Do it right, not fast.'
This is where HP Nio comes in. Nio is trained on a decade of HP data and continually fed with knowledge from service engineers and customers. It is built to the same enterprise standards for security, privacy and ethics that HP applies as an IT company. Diana says adoption has exceeded expectations, with around 70% of responses getting a thumbs up. Every thumbs down is analysed and fixed. Nio is updated daily, with major releases every three to four months, and support for more languages arrived in September.
What has surprised HP is how people are using it. Most questions are still 'how-to' queries, used much as you would a search engine. Even so, HP is seeing higher uptime on presses where operators use Nio, and fewer calls to support, so HP now plans to encourage more ambitious uses, such as asking Nio for a two-week plan to raise output.
Nio is due to go into beta in Site Flow in the coming months. It is also coming to HP's colour and quality tools, colour being 'the most complex topic in this industry', according to Diana, and one she admits she could talk about for hours.
How does all that sit beside HP's talk of autonomous production? For Diana, the two go hand in hand. 'It's not that there are no humans involved,' she said. 'It's that the humans that are still there are super-powered, super-equipped with all the AI tools around them.' Instead of moving boxes and clicking buttons, people make the final call on decisions that have been prepared for them: a press is down, a job will be late, here is a suggested reschedule, and the person simply approves it or not.
Diana believes the time saved should go into growing the business. Costs come down because production is more streamlined, and revenue can grow without adding cost. She compares it with how many of us now use AI in the office: less time lining up boxes on a slide, more time on the message itself.
“Take one, make it a success. Take another, make it a success”
Where should a 25-person print business with three systems, a handful of spreadsheets and one indispensable employee begin? Diana would start with education. Start by understanding why you want to digitalise. It might be order intake, selling online, diversifying, cutting costs, better customer service or faster production. Then map your bottlenecks, set priorities and look at how others have done it. She prefers 'digitalisation' to 'automation', because it covers the business as well as the production floor.
'Take one, make it a success. Take another, make it a success,' she said. If you want to move faster, ask for help. For smaller businesses, HP points to PrintOS Production Hub as a first step away from paper and Post-it notes, with Site Flow for more advanced automation.
IN DIANA’s OWN WORDS
Why use Nio rather than plugging a public AI tool into your systems?
'It's not only that it's trained on ten years of data, hyper-specialised, with a ton of knowledge. It's also private, it's also secure, and it's also ethical.'
Where will AI make the biggest difference first?
'Analytics, insight, scheduling, image enhancement, order workflows, creating workflows across the factory. AI is going to significantly change how we work in the next few years.'
What should a smaller print business do first?
'It begins with education. Why do you want to digitalise? What do you want to get out of it? Then break it into pieces and make priorities, like with every strategy.'
Looking ahead to 2030, Diana believes the industry is underestimating the power of AI agents. 'I wasn't calling agents a year ago, and I have agents working for me now,' she said. Her vision is of 'agents talking to agents', managing much of the ecosystem, but there is no shortcut. 'To get to agentic, first you need to get to AI. To get to AI, you need to get to digital. To get to digital, you need to get to data and automation.'
For her own team, the challenge is keeping pace. 'In the years we are living in, there is a risk that from one month to another, the product we are working on feels five years old,' she said. That sometimes means being brave enough to stop a project when a better approach appears. 'Learning, learning, learning. Stay ahead, stay ahead, stay ahead.'
For print businesses, the press is only part of the story. Those that thrive over the next few years will be the ones that tackle the friction around it and take a careful, practical approach to AI.
Listen to the full conversation with Diana Pascual on the FuturePrint podcast.