Brightpick is a robotics company that develops autonomous mobile picking robots powered by AI to enable lights-out warehouse operations and improve efficiency.
Why do so many businesses struggle to achieve returns on their AI investments, and what tangible results are your robots delivering for customers?
In many cases, the problem is not AI itself but whether people know what to build and how to tie it to real financial impact. A lot of people are empowered by AI, but that does not mean they understand how to create a product with real ROI.
In our case, the value is very concrete. Warehouses still involve a huge amount of repetitive manual work, and our robots improve efficiency and, in some cases, replace roles that companies already struggle to fill.
The benefits are immediate. In a CapEx model, where the customer buys the system, we often see a two-year ROI. In a robotics-as-a-service model, customers can cut operating costs by half as soon as the month after deployment. The implementation may take around three months, including API integration, but from month four the savings are already visible. In pharma, for example, 99% of objects are picked by AI, but ultimately customers care about ROI. If the business case is proven, the project succeeds.
You have stated that early adopters of AI are often the ones who fail. Why?
Yes, that may sound counterintuitive, but I do think it is often true. Many early adopters move before the technology and the business case are mature enough. They adopt too soon, results fall short, and then they blame AI, when really the issue is timing and a weak commercial case. I remember going to Silicon Valley 15 years ago and being shocked that the internet was much slower than in Slovakia. The explanation was simple: they adopted earlier, but the quality was worse.
I think something similar is happening with AI. If you compare the models people were using two years ago with the best systems today, the difference is obvious. Early adopters sometimes fail to realize they are experimenting with technology that is still immature, and they do not always understand the business case well enough. So I would not call it an AI mistake. I would call it a business case mistake, and sometimes simply the mistake of adopting too early.
There is a lot of competition in warehouse automation. What makes Brightpick unique, and how does your new robot, the Gridpicker, perform?
When people in tech see our robot, they often think it’s just sexy. It has a robotic hand, a 3D eye, and visible AI. But that is not what actually makes it special. What is unique is that it is an autonomous mobile picker. Most other systems still work in a back-and-forth model: a robot brings a tote to a person, the person picks one item, and then the tote goes back.
Our robot goes into the warehouse and collects multiple items directly, more like how a person would naturally work. If you are cooking, you do not go back and forth ten times for ten ingredients. You collect them together. That is why our system is fundamentally different and, in our view, two to two and a half times more efficient than any other system. That is also why the new product is very price competitive and why we can deliver very high labor savings.
How significant can labor savings be in practice?
They can be very dramatic. We have seen warehouses that had 50 pickers go down to three. In some cases, we can run the warehouse fully lights out at night, with no people on site and only the robots moving. That creates another major advantage, because the automated overnight shift prepares all the robot-pickable items before the morning.
If there are still some very complex items that the robot cannot handle, those can be picked later by people. So instead of running multiple labor-heavy shifts, customers can move a large part of the work into a fully automated process. That creates a major boost in labor efficiency and throughput.
How widespread is warehouse automation today, especially among smaller companies that may not have the capital to invest?
If you look at large companies, maybe Fortune 500 or Fortune 1000 types, many of them are already around 20 to 30 percent automated in terms of warehouses or processes. So they are fairly advanced, although there is still room to go further. Smaller warehouses are different. They often do not have the capital, and they also do not have the people with the IT capability to connect and manage these systems.
That is exactly why we are pushing robotics as a service. It lowers the barrier. Instead of asking customers for a major capital purchase, we can tell them to take the operating cost they already spend on labor, cut it in half, and use that to fund a nearly fully automated warehouse. I think that is the next wave of automation. Large players are already coming. The bigger challenge, and opportunity, is bringing automation to smaller operators through the right product and business model.
Do you see Brightpick expanding further into manufacturing?
Yes, we do see traction from manufacturing. My previous startup was deeply involved in automotive and manufacturing, and we sold it to Zebra Technologies a year ago. Brightpick is in a way our second startup, so I know that world well. Today, maybe 15 to 20 percent of our leads already come from manufacturing, and I think that could grow.
The connection is simple: whenever you manufacture something, logistics is involved. You still need to move and supply components, so there is a natural overlap with what we do. That said, our primary focus remains e-grocery, pharma, small electronics, drugstore items, and beauty products, especially categories where goods are stored in totes and picked in volume. That is where our system performs best.
Can you talk about your recent partnership with NAPA Auto Parts and what it means for Brightpick’s future?
For us, it is a major milestone. We have worked with listed companies before, but this is still exceptional. It involves over 100 robots per warehouse, which is a huge fleet, and the warehouse itself is enormous. For any startup, there is a kind of stair theory: you climb step by step, building trust with bigger and bigger customers. Once you reach a certain point, it becomes much easier to talk to anyone in the market and say that you already work with NAPA.
What I also like is that they have a strong mindset and a dedicated budget. It looks like a serious, well-structured plan. That tells me big companies are already fully committed to automation. You no longer need to convince them that automation matters. Now the discussion is about ease of operation, visibility, and data. Those are much more advanced conversations, and that is why I am excited about growth.
What are your top priorities for the coming year, and where do you see the biggest untapped opportunities?
One big priority is our own manufacturing, because we are moving toward producing thousands of robots, which is a completely different level. Manufacturing is not rocket science, but it is still very difficult, and scaling it well is essential. Beyond that, we see opportunity in expanding the number of warehouse processes we automate. Picking is only one part. There is also packing, replenishment handling, and other connected tasks.
The more of those processes we automate together, the better the outcome for the customer. We also see future potential in what we call semi-humanoids. That does not necessarily mean a robot with a full human form, but a stripped-down system with two hands designed purely for efficiency. The latest models are getting close, though I would still say they are not quite primetime ready. Over the next one or two years, I think that will open new segments for us.
