Summary of Cio Rom Kosla de The importance of emerging technology At Hewlett Packard Enterprise (HPE) probably resonates with any senior executive: “IA is in our mind. “”
Research suggests that Kosla is far from alone. More than Three -quarters (78%) of business leaders According to a recent McKinsey study, their organization uses AI in at least one commercial function.
Also: 4 ways that your organization can adapt and prosper in the AI era
Kosla told ZDNET that HPE uses third -party applications with Integrated AI capacities And has spent the last 18 months to develop an internal cat solution called Chathpe, a generative AI center used for internal processes.
Here are four ways to use Kosla’s experiences to transform AI into a commercial advantage.
1. Establish an AI strategy
Like other first -rate companies, HPE uses several external AI models and technological platforms. Professionals use Microsoft Copilot To increase productivity, while developers use Github co -pilot.
“The objective here is that, while we deploy licenses to individuals, they exploit these technologies to make their daily jobs,” said Kosla.
The strategy becomes more interesting around Chathpe, a tailor-made platform that is fueled by Azure and Openai technology.
Also: This free Google tool transforms AI into your research assistant
Kosla said that the company has a user case pipeline that are assessed and assessed. For example, a legal team may want to use AI to examine contracts, extract specific information and generate new models.
“What we do next is to work with them, to take advantage of Chathpe and point them to specific source data,” said Kosla. “We will also limit access to data, because we do not want people to use information without care and consideration.”
He also explained how use cases could bubble in other areas of the company.
“In the case of operations, it is more around cat. You can consider it as something for customer service. For example, there could be a use case where people want to find a part number. They use Chathpe to access information, and technology gives an overview to help customers,” Kosla said.
He added that marketing can use Chathpe to identify what campaigns could be reused, while finance could opt for analysis use cases.
2. Train everyone to use AI
Kosla said that her staff often discussed the concept of majors and minors in their trip to become a mature computer organization.
“If you think of a university, you may be focused on the economy, and that’s really what you want to do,” he said. “However, it does not hurt to learn finance, marketing or making a minor in computer science. All these things create a healthy vision of the opportunities that you can apply.”
Also: Is your business ready for AI? 5 ways of avoiding late
According to Kosla, this is the approach that Hpe adopts at AI. The company hopes that all employees consider and competent in various majors and minors as areas of competence.
“If your adult is in supply chain, you need minor in AI. Or if your Major is in AI, you have to minor in supply chain,” he said. “You must be as diverse as possible in your knowledge, because everything we do will cross the limits of other areas of the company.”
Kosla explained how the concept of HPE of majors and minors is reinforced by IA IA learning and development. “We allow this training with our HR systems. People can come and learn. The goal here is to put AI training available to everyone,” he said.
But this orientation does not stop at training – Kosla underlines the importance of follow -up after the fact. “As they adopt technology, my rule is always:” If you learn, you need to use it. “It is not great to have a certificate and not to be able to apply knowledge.”
3. Put the railings in place
Kosla has said that it is already clear that AI will permeate each organization, function and role, which leads to a proportional impact on access to information. Intelligent business leaders will focus on managing the relationship between agent technologies and their human counterparts.
“From a computer point of view, I think of checks and railings,” he said. “You need the human factor in understanding questions like:” What is the machine and what decisions do they make in your name? How do you make sure that these guards are modified, and tightened or widened, depending on the scale of what you are trying to do? “”
Also: The 20 best IA tools of 2025 – and the thing n ° 1 to remember when you use them
Kosla said these Railing can only be established And refined by the IT partnership effectively with the company.
“A large part of the work was initially around questions such as:” How do we make sure that we manage rapid engineering correctly, that the data extracted is limited, the quantity of training is managed correctly and that the engine responses are accurate? “”
He continued: “There are key sentences to which we do not want our engines to respond because, potentially, the answers could be dangerous or harmful. So, what we are trying to do is guarantee that the incitement is for the use of companies, but also that work does not generate too much request, because we want to avoid high costs.”
4. Consider where you will then go
Kosla said that HPE’s internal approach to AI continues to mature and that the company will mix internal and external models depending on the requirements. The current company’s emphasis is on Use of AI to increase the operational activities of staff.
However, Kosla also recognized that AI agents evolve quickly. His company already uses software from other suppliers, such as SAP and Salesforce, which are implementing agent services.
Also: 10 OPENAI strategies uses to create powerful AI agents – which you must also use
One of the key questions HPE considers is the orchestration of AI through its exploitation model from start to finish.
“If we orchestrate between two suppliers and they have two different agents, but an individual uses both and they rotate their chair between one request and another, they could see that they are not talking about it,” he said. “So we have to build this orchestration layer, or we have to find best practices around this problem.”
Kosla said these technical problems will temperate part of Media threshing around agents.
“When you think of an individual, they are not linked to a single application,” he said. “And this is the challenge we face. There is a lot of sales regarding agents. But you are not an agent that only finances – you also make finance, quotation, pricing and all these systems interact together.”