How to measure the value of AI projects by Craft Heinz

Companies are polishing approaches to identify the correct AI use cases and to show the business value that comes together.

For Craft Heinz, the success of the initiative is measured by a mix of quantitative and qualitative evaluation. PAT NESTOR, decision -making intelligence product and platform officer In food and beverage giants.

Nestor said in the CIO DIVE, “It is a place where business case evaluation begins in terms of quantitative aspects. The team should be able to connect the use cases with a wide range of business goals, such as sales promotion or cost reduction.

Quantitative evaluation is paired with qualitative values ​​such as its influence. Pure promoter score And the level of adoption.

Nestor said, “We provide a sound feedback loop that can measure and track NP, adoption and participation scores to be refined and evolved.

More C-Suite, Investor and Board of Directors, the AI, which focuses on AI and supports efforts, focuses on the larger budget, and the technical leader should be able to get the formula properly by selecting an expandable and beneficial use case. Trace success after placement.

AI project failure rate has increased This year, a company is graphple Inhibitory battery. Almost all -97% -oThe F Business struggled to show the business value of the generated AI according to the InformaticA report released in February.

Successful company Credit use cases are aligned with a wide range of strategic goals for supporting project priorities and considering available resources and predictable schedules.

Nestor said, “One of the most important things for us at my favorite point is to see AI, especially AI as a solution itself, but rather an important factor that can overlook bigger programs.

There is a good effort to press the test use case before Kraft Heinz selects a location for a large AI bet. This is also a stage in which success criteria are established and models are created.

Nestor said the team first tried to answer the next question.

• Where do the company want to go?
• How can you evolve potentially?
• How is it measured?
• How can I evaluate the progress along the way?

Nestor said, “A high -end -rich funnel is very important to have versatile ideas and versatile business cases before actually crying.

But according to Nestor, the failed experiments or pilot phases are not the end of the world. Learning from failure is part of all innovation processes.

Nestor said, “The bottom line is that we must expect and accept some of the levels of failure.

Dip to AI

Earlier this year, Craft Heinz reconstructed the technology governance structure. The company has a global technology umbrella and a global technology umbrella Company’s Global CIO, Corrado Azzarita.

As part of restructuring, Nestor moved from Kraft Heinz’s global product manager to lead the newly formed Decision Intelligence Products and platform teams.

Nestor said, “Basically, it is our data for the AI ​​Value Chain organization, so there are all aspects and sub -teams, part of the decision -making intelligence portfolio. The team also manages the platform and product in the domain.

Reorg is consistent with a wider strategy change as the company gives up for six consecutive quarters of the total reduction. Food dive has been reported transmission. Craft Heinz Efficiency unlock and productivity improvement.

To achieve this goal, KRAFT Heinz is a basic platform of Khai, a manufacturing and snowflake of Data Management Infrastructure, depending on strategic technical partners, including Microsoft, including Microsoft.

According to Nestor, Khai took about 180 days from the idea and has about 13,000 team members worldwide. The company wanted to democratize the creation AI access without sacrificing security. The AI ​​engine is made internally and is rescued to provide a “one -way door”.

Nestor said, “The rich information that Khai is trained and the basic LLM has been trained still can still come in.

Team members use the platform to analyze and summarize documents or dense industry reports, and optimize and access standard operation procedures in the factory floor. The engine also helped to simplify China’s SAP rollout and automate everyday finance. Nestor said the platform is bookmarked in the browser and uses it every day.

“This is actually an essential part of a variety of workflows.” Nestor added that the number of active users is expected to increase.

The company expects the ability to mention. A colloquial body like KRAFTGPT It is completely integrated into Kai. This tool is now called PlantChat and is used to help strategic decisions in the manufacturing ecosystem.

Kraft Heinz has several different AI pursuit. Supply chainIncluding it Pickle production process.

Nestor said, “We think that each of the craft Heinz employees has quantitative and qualitative value.