by Cassandra Balentine
Artificial intelligence (AI) has entered the conversation. However, whether it’s resistance to change or fear of the unknown, many are reluctant to trust AI.
“I sent a broadcast email out to our customer base to ask if anyone had any suggestions about where AI in Print Shop Manager would make their life easier,” recalls Rich Giles, president, AACRO Computer Systems, Inc. “I was surprised at the response. The largest number of responses indicated they did not want AI in the program. They do not trust it.”
Nevertheless, AI is making its way into processes across the board. It may be time to consider how AI can best help your business processes and fuel automation.
After all, Ginni Rometty, former CEO, IBM, may have said it best—“AI will not replace humans, but those who use AI will replace those who don’t.”
Automation Considerations
Automation—especially AI-driven—is poised to transform print workflows tools including management information systems (MIS), enterprise resource planning (ERP) solutions, and end-to-end workflow offerings.
“The push for automation is never ending,” asserts Sashikanth Goparaju, CTO, InSoft Automation. “The industry, its machinery, and way of doing business are changing faster than ever before and automation must keep pace with these changes. The digital world is all around us, and we all expect instant answers to our questions.”
Joseph Lehn, director of product management, PressWise, feels that the shift to automation is more of a steady movement versus a trend. “Everyone needs to do more with less, which means finding the most efficient ways to get things done. Run lengths are often shortening as the focus of the audience becomes more concise. Short-run custom jobs are becoming more common.”
Rules-Based Automation
Rules-based automation is one of the most basic forms of machine intelligence. It is used to apply directives that store, sort, and manipulate data based on predefined perimeters, generally set by humans. Estimation tools often utilize rule-based technology to drive automation.
“We have seen a significant shift in the past few years that continues to accelerate to rules-based automation or even AI-driven automation,” says Nick Benkovich, chief product officer, Print ePS.
He predicts this will continue to accelerate as service providers are challenged with hiring skilled workers and managing the volume and complexity of the work they need to deliver.
Driven by the demand for speed, consistency, scalability, and profitability, rules-based automation not only improves responsiveness and customer experience—it also unlocks workflow automation downstream. With rules-based automation in estimating, accepted quotes can automatically convert into production jobs, trigger inventory checks, notify purchasing, and populate dashboards and reports, forming the backbone of an end-to-end automated MIS. “This level of integration isn’t just a trend, it’s becoming the baseline expectation in a modern print operation,” he asserts.
Jef Stoffels, head of marketing, Dataline, also sees a growing trend towards automation and rules-based estimation, particularly in segments of the market where order intake is increasingly digital. “Print companies that rely heavily on web to print (W2P) or print on demand platforms are especially keen to streamline their workflows. In these environments, standardized products and repeatable processes create ideal conditions for applying estimation rules.”
Helle Vogt Mikkelsen, head of marketing and strategic partnerships, PrintVis, agrees, noting that fast and accurate estimation is admittedly a major challenge for many print providers, and automation—even AI—can help. For example, she says PrintVis is investing heavily in AI-enhanced estimation, which is expected to launch soon. “Our system already offers a flexible setup that allows users to create templates and estimate jobs quickly, automatically, and consistently to ensure that no details are overlooked. By guiding estimators through all relevant factors, PrintVis helps reduce errors and ensures accurate pricing every time.”
Allison Taylor of Franklin Estimating sees a trend towards automation and rules-based automation for common printing projects. However, for special projects it is not as likely as specs on the project, although similar, many have too many factors that change. “An example is cost for stock, inks, and even staff.”
AI and MIS
The integration of AI and machine learning (ML) in MIS and similar systems also plays an increased role in the future of automated estimating.
Craig Powell, GM, North America, Print IQ, feels that AI and ML are just beginning to play a transformative role in estimating within print MIS platforms—moving beyond traditional rule-based logic to unlock predictive insights, intelligent recommendations, and continuous optimization. “While the print industry has historically relied on static estimating rules, AI introduces the ability to learn from past quoting and production data. For example, an ML model can analyze historical win/loss rates, actual versus estimated job costs, and material usage trends to recommend pricing strategies or highlight inefficiencies in estimating logic. Over time, these systems can adapt to evolving customer behavior, seasonal patterns, or supplier changes—delivering smarter, more competitive estimates.”
By analyzing historical data, job specifications, and customer behavior, Stoffels says AI-driven systems can suggest optimized pricing in real time, even for complex or customized orders.
“AI and ML are definitely transforming the way estimations are handled within MIS environments,” offers Stoffels. “On one hand, they help estimation professionals work significantly faster by automating repetitive tasks and learning from past data. On the other, they bring a new level of precision and flexibility to pricing structures, particularly in web to print and print on demand workflows—both business to business and business to consumer.”
“The complexity of the work, the shortage of skills, the less sophisticated buyers, and the need to turn more estimates faster have driven the need to use tools like Natural Language Processing for buyers, sales representatives, and even customer service to become part of the estimating process simply by describing what that final product needs to be and allowing AI to do the work,” offers Benkovich.
While some ML techniques have been applied to estimation for more than a decade now, Goparaju says there is significant success in certain domains and not so much in some. “With respect to cost estimation for printing, the results in my opinion are not up to the mark,” he cautions.
When dealing with automation, accuracy is critical. Gerald Clement, partner, Computer Productivity Services Inc., explains that its estimating application is very structured with many automatic checks and balances. “Users ask for system changes at the plant level to password protect non-standard changes. Users can also create their own system alerts for any condition. In other words the system is very good at catching issues but people still have to decide what they want to do about it.”
Overall, the goal isn’t to replace human estimators, but to empower them with intelligent tools that accelerate decision making and reduce errors. As AI models mature and data volume grows, Powell expects estimating modules within MIS platforms to evolve into learning systems—continually refining the accuracy, profitability, and speed of every quote generated.
Taylor points out that while AI is expected be found in many MIS solutions as a selling point; from a programming standpoint, it has little to offer right now to serious developers. “As AI scheduling and inventory management can offer time and cost savings, taking responsibilities from users is risky business.”
Benefits and Limitations
The primary benefits of AI in MIS include time savings, improved resource allocation, increased accuracy, faster estimating cycles, proactive error detection, and continuous optimization.
“With AI, users can automate time-consuming tasks, instantly summarize complex order data, receive real-time decision support, predict needs, and suggest next steps,” comments Vogt Mikkelsen.
Further, Powell says ML can assess past quote performance and recommend pricing strategies that align with customer behavior or market conditions, helping boost quote conversions. AI can also suggest optimal configurations or pre-populate estimates based on job type, past behavior, or seasonal trends—reducing estimator workload and response time.
AI and ML offer the potential to make estimating within a print MIS more powerful. “One of the biggest advantages is the ability to tap into the vast amount of historical data already stored in the MIS—every estimate, job detail, and outcome becomes valuable input for smarter, data-driven decisions,” offers Stoffels.
“AI enables adaptive learning—quote accuracy improves over time as the system digests more real-world outcomes, feeding back into pricing models and workflow decisions,” adds Powell.
Like any AI application, Vogt Mikkelsen says its effectiveness depends on the quality of the data and the consistency of usage patterns. “AI doesn’t replace human judgment for highly customized or unusual jobs—but it provides invaluable support.”
Stoffels agrees, adding that AI and ML require clean, structured, and sufficiently rich data to deliver reliable results. “If the data quality is poor or inconsistent, the output will reflect that.”
Unlike rules-based automation engines, Powell points out that AI/ML models can be opaque. “This lack of transparency may make users wary of trusting recommendations without clear rationale or audit trails.”
Goparaju says that real-world mathematics of cost estimation for print products is non linear. “Sometimes, just a millimeter change in the dimension of a product could cause a significant difference in the cost, as the fit of the product on the print layout could change.”
Therefore, Goparaju suggests that training AI/ML models for such scenarios is difficult, if not impossible. “Further, the model can fail when challenged with jobs outside the training data, directly leading to financial losses. In financial matters, the emphasis on precision as well as accuracy is much higher. To generate confidence in the quote they send out, users generally want to look at the actual impositions/layouts based on which cost estimates are made. However, that is not possible when using AI/ML technics.”
Powell admits that AI performs best in high-volume, repeatable contexts. “Uncommon or highly customized jobs still require human judgment and can’t be reliably estimated by pattern recognition alone.”
Further, embedding AI/ML into a live MIS environment requires thoughtful architecture, real-time data access, and sometimes new infrastructure—making it a longer term investment, adds Powell.
Overall, Powell concludes that AI/ML within a print MIS offers tremendous upside—but only when paired with quality data, thoughtful implementation, and human oversight. “The future lies in hybrid intelligence—where AI accelerates insight and consistency, and humans guide the edge cases and strategic decisions.”
Taylor feels that taking responsibilities from users of software and placing it mostly in the hands of ML can certainly miss special circumstances. “However, if implemented well, with user input throughout the AI’s functions, those risks can be reused to save time and costs.”
Automated Estimating in MIS
Automated estimating is increasingly tied directly to print MIS. This alignment is both logical and necessary for modern print operations.
The lifecycle of a job starts with a customer query and ends with delivery of a finished product. “An important event in this lifecycle is the customer accepting a quote and confirming an order. It is only then that the job hits the production line. The essential task of an MIS is managing the job after the order confirmation. The quotation cycles before the order confirmation can be considered as a welcome add-on module to an MIS for a much smoother and more integrated system. But if the estimation module of the MIS is not accurate with its numbers, then it makes a lot of sense to have independent, more accurate estimation software,” adds Goparaju.
Estimating is a key function of any print MIS, and automation only strengthens its value. “If AI and ML can improve the speed, accuracy, and success rate of estimates, then these capabilities become key assets within the MIS,” says Stoffels. “Importantly, once an estimate is approved, the same data should flow seamlessly through production, logistics, shipping, and finance. This end-to-end consistency is only possible if the automated estimating function is fully integrated within the MIS, making that connection not just beneficial, but essential for modern print companies.”
Powell says integrating automated estimating within an MIS makes sense for many reasons. One being centralized data. “Estimating relies on accurate costs, inventory, production capacity, and pricing rules—all of which are managed within the MIS. Tying estimating to the MIS ensures that quotes are grounded in live, reliable data.”
Also, it helps to automate workflow. “When estimating is embedded in the MIS, accepted quotes can instantly trigger downstream actions—like job creation, stock reservations, purchase orders, or scheduling. This creates a seamless, touch-less workflow from quote to delivery.”
Automated estimating within an MIS also brings margin and cost control, self-service and portals, as well as scalability, according to Powell.
Powell points out that standalone estimating tools often suffer from data silos, duplicate entry, and disjointed workflows. “By contrast, tying estimating into the MIS enables a single source of truth—which is critical for accuracy, efficiency, and business intelligence.
printIQ has fully embraced this model. “Our estimating engine is not a bolt-on; it’s core to the system architecture, integrated with inventory, production, shipping, outsourcing, and invoicing. That’s what makes end-to-end automation possible—and why we believe estimating belongs at the heart of the MIS.”
Lehn argues that automated estimating is not new to print MIS, as the PressWise system has combined the power of the estimating engine with the storefront, the estimating system, and job creation to ensure consistency of pricing regardless of how the price was requested from its inception. “On short-run digital jobs, the margin is often small enough that creating the need for a person to review the order before it goes to print negates the margin. Automating the pricing that is always consistent makes these jobs possible.”
Goparaju shares that InSoft’s Imp layout planning software is used by many MIS to improve the accuracy of their estimation modules. The benefit of using such MIS is that they base their cost estimation on precise layouts calculated by a smart layout planning tool. “Yet, there are many printers who have MIS that do not have such an estimation module. For them, InSoft offers Imp+, a smart and highly adaptable estimation tool. Almost all of our customers who use this tool, export an XML file from our software that is later imported into their MIS to create a new order in their production environment.”
Accurate and Automated
As the print industry moves to shorter run, tighter turnaround work, the need for accurate, flexible, and intelligent estimating is apparent. When properly implemented, rules-based intelligence and AI enables estimating engines within MIS and similar systems to automatically produce accurate quotes. However, it is imperative to ensure that solid, accurate data feeds into the system and that many checks and balances are in place to reduce the risk of under quoting.
For more discussion on this topic, check out our recent webinar, Automation in Print Management.
Sep2025, DPS Magazine
