AI Panel (CONT’D FROM PAGE 22)
We are probably looking at a $50,000 investment and a one- to two-year timeline. The bigger cost is time. The longer we wait, the longer it will take. Beyond the price tag on the investment, the bigger question is whether we can pull people away from what they are currently doing or whether we need to add staff. It becomes a risk assess- ment: is the effort worth it, and are we confident it will be successful? Alvarez: One of the biggest decisions is which platform you are going to use. Many people started with the Copilot experience, and even Microsoft seems to recognize that its strength is Office and organization. Now OpenAI mod- els are becoming more accessible, Claude tools are avail- able, and Gemini may also become part of the ecosystem. For adoption, especially with bigger solutions, work- flow matters. Decide on the platform, stay disciplined, cre- ate a governance workflow, and bring solutions in slowly. Otherwise, you may end up subscribing to 10, 15, or 20 different tools, and that gets expensive quickly. Your tool of choice will probably live within your office platform. I have asked people, “What is the number one application in your enterprise?” They may say it is Kiwiplan or Amtech, but Outlook is open all day. People live in Of- fice, Google Sheets, email, and Teams. That is where most communication happens. Zlatic: Let’s move into security. What precautions should companies take to protect customer data, pricing,
point where we do not need to hire additional people. I do not view it as reducing staff. It’s more about making the people we already have more effective. Blizzard: Going back to the shipping example, it costs about $20 a month for that manager to use the tool. The freight savings from the first time we ran the report cov- ered roughly two years of that cost, not including the time savings. Is it the most quantifiable ROI? No. I think the in- dustry is still struggling with how to prove ROI for AI invest- ments. But in that example, the value was very clear. Zlatic: Let’s talk about time to deploy. These quick wins are good, but what does time and investment look like for some of these more in-depth solutions? Blizzard: Gathering our data took the most time. We are in year three. The first year was data assessment. The second year was building out the platform. The third year is operationalizing it. That is the data side. But quick wins are much faster. They can be as simple as signing up for a ChatGPT license, developing an AI policy, and training users on the tool. In my earlier example, the training took about 30 minutes. Perkins: We are currently working on an ERP schedul- ing system transition. One use case we are considering is automatically reading customer POs, then looking across our fleet of machines to determine how quickly and effi- ciently we can produce the order.
CONTINUED ON PAGE 26
Serving the North American Corrugated and Folding Carton Industries Since 1985 DRIVE your message to the GREEN... with the credibility of Board Converting News behind it! Sponsor a weekly Monday morning issue blast or we can ‘Tee It Up’ any time you like to our full online circulation of 7,500+. You provide the copy, images and link to a video or website and we’ll do the rest!
Contact Len Prazych 518-366-9017 lprazych@nvpublications.com
24 June 29, 2026
www.boardconvertingnews.com
Made with FlippingBook interactive PDF creator