HP Vietnam General Director: To Unlock AI’s Value, Businesses Must Clear Bottlenecks Before Speeding Up

AI adoption has never moved faster among Vietnamese businesses. Yet productivity has not made a comparable leap. Part of the reason may lie in how companies are bringing AI on board in the first place. Most are still running on processes built for a pre-AI world, then simply bolting AI onto them without ever rethinking how the work actually flows. The result: they're moving faster, but running headfirst into the exact same bottlenecks as before.

It's exactly this problem that Nguyen Minh Duc, General Director of HP Vietnam, dug into with host Tu An on Vietnam Innovators. With AI now reaching into every corner of business, the question isn't whether to use it anymore - it's how to use it in a way that actually creates value.

Speeding Up Isn't The Same As Creating Value

Mr. Duc has noticed that the average employee now runs anywhere from 3 to 10 AI Agents a day - a number that captures how a lot of people approach AI: stack on more tools, and productivity should follow automatically. But neither the sheer number of Agents nor how convenient each one is explains why some companies break through while most don't.

One story stuck with him. A major HP dealer in Vietnam grew revenue by more than 60% in a single year, with headcount barely moving. What made the difference wasn't stuffing AI into the old process, it was rebuilding how the business ran first, then letting AI models automate what had already been reorganized.

As he puts it, the question was never who uses AI more, it's who uses it to create more value. But once a business reorganizes its operations and lets AI reach deeper into internal data, a different kind of friction shows up right away. It's called data security.

As AI Digs Deeper Into Data, Security Becomes The New "Friction"

AI is spreading so rapidly that businesses are being pulled into a new race: adapt proactively or risk falling behind. More companies are therefore encouraging employees to use AI to boost productivity. But awareness around data protection has not kept pace with adoption, raising concerns over the leakage of sensitive information.

Cloud-based AI models typically need to send data to centralized servers for processing. For highly sensitive industries such as banking and personal data management, that extra step can introduce a risk they simply cannot afford to take.

This is where HP AI Box comes in. The solution brings AI processing locally through a range of devices designed for different business needs, from small and medium-sized enterprises to large corporations. Notably, AI Box can now run large-scale Frontier Models, comparable to models such as ChatGPT, directly within a company’s internal infrastructure. This allows businesses to build a closed-loop data processing environment without sending sensitive information outside their systems.

The same locally approach is also evident in Scan AI, a solution co-developed by HP with several businesses. In practice, companies process countless documents containing sensitive information every day, from personal records to internal contracts. For this kind of data, even passing through an intermediate Cloud layer can introduce an unnecessary point of exposure.

Running Scan AI on the AI Box platform keeps the entire scanning and processing workflow within the company’s internal infrastructure. Data does not have to leave the system at any stage, while employees can still perform AI-powered tasks locally.

To address this layer of risk, HP equips its enterprise computers and devices with Wolf Security, a security solution designed to protect users at the endpoint. The system can automatically assess whether a link is safe before an employee opens it and cut off the connection when it detects a threat. Each HP computer also comes with a dedicated Wolf Security chip that operates independently and can automatically disconnect and shut down the device when it detects an external security risk.

In other words, HP is taking a holistic approach to enterprise AI security: controlling the flow of data through HP IQ and Scan AI while quietly guarding everyday actions - from clicking and opening to sharing - with Wolf Security as a protective layer.

But AI technology can only create lasting value if it also makes everyday work feel more effortless, including at the seemingly smallest points of interaction: turning on a computer, joining a meeting, or printing a document.

People Remain At The Center Of Every Automated Workflow

Mr. Duc has zeroed in on the three touchpoints employees interact with most at the office: the computer, the printer, and the meeting room. Getting ready for a meeting, he says, usually eats up about five minutes just booting up, hunting for the link, and connecting to the screen. Nobody calls that a big deal - but it's exactly the kind of thing that can be automated away to make the workday smoother. That's why HP is building its products to connect all three seamlessly, stripping out what he calls "friction" in everyday work.

Instead of typing in a link and password every time an employee walks into a meeting room, the room's camera now recognizes them and connects their laptop to the call automatically - cutting a setup that used to take 5 to 10 minutes down to almost nothing. The printer works the same way, connecting itself the moment it recognizes the right person standing nearby.

In its next generation of laptops, HP is rolling out HP IQ, which processes data right on the device without needing an internet connection or a middleman server. HP IQ also lets two computers recognize each other and share data directly.

All of these small pieces fold into what HP calls "For Your Best Work." The idea was never to let AI take over for people - it's to clear away the repetitive stuff, so people get their time back for what machines still can't do: creativity, and talking to each other.

For Mr. Duc, making AI actually count starts with leadership. Before picking any solution, leaders need to review and rebuild their own operating processes first, then go looking for tools that fit their scale and their team's skill level. "Change has to start at the top for a business to move fast and go far," he says.

Friction in how a business runs doesn't disappear just because everything's moving faster. It only really goes away once a company restructures the right things and builds infrastructure secure enough to let AI go deep into its data without creating new risk. That's exactly what HP is doing for itself, before bringing the same approach to businesses across Vietnam. As he puts it: "AI isn't the destination — it's the process." And in that process, every HP device is pointed at one thing: For Your Best Work.