A Woman from Kazakhstan Figured Out How to Control Drones with the Power of Thought. Her Startup Is Already Being Tested in the USA

Dara Tumenbayeva is originally from Almaty. For several years, the Kazakhstani entrepreneur built her career abroad, working on and launching her own AI projects. This year, she founded a startup called Black Swan. Her team is developing a device that makes it possible to control drones using brain signals and hand movements. The product is already being used in the event industry, but the bigger goal is to break into US government defense contracts.

For the joint Digital Business and Astana Hub project «100 Startup Stories of Central Asia», Dara talked about why she switched from medicine to DefenceTech, how the neuro-helmet actually reads brain signals, and why her startup stages spectacular drone races.

«Ignoring conflicts and pretending they don’t exist is childish»

– Can you tell us what you were doing before launching Black Swan?

– I built my career as a software engineer. I studied in Malaysia, then worked across Asia and Europe, including at Amazon in Spain. I also launched a few projects of my own. One of them was a startup in dermatology and aesthetic medicine called Doro AI. We worked with clinics in the US, Latin America, and Australia.

Back then, we built a platform that let people see what they would look like after cosmetic procedures. Our typical client was someone who had already had a lot of Botox and was still looking for new ways to improve their appearance.

At some point, I realized this project just wasn’t my reason to get up in the morning. It had traction, clients, and revenue. But I didn’t feel like I was doing something that truly mattered. I’m the daughter of a doctor, and it was always important to me to bring real value to people. With Doro AI, the focus was on external changes, on appearance. And that didn’t feel meaningful enough to me anymore.

That’s when I started thinking seriously about where to go next and exploring different directions.

– What were you looking for in your next project?

– I was interested in technologies that help better understand a person’s condition. Things like levels of focus, stress, and attention. A lot of that interest came from personal experience. I went through depression and panic attacks and started digging into how the brain actually works. That’s when I began looking into non-invasive neurotechnologies, specifically brain-computer interfaces. These are devices that read brain signals externally, without surgery or any kind of physical intervention.

Put simply, you wear a special device with sensors on your head. It doesn’t literally «read thoughts». Instead, it detects what state a person is in: focused, tense, relaxed, or distracted.

This data can be interpreted and used in practice, for example to control devices without buttons or joysticks. From the outside, it looks like movement powered by thought. In reality, it’s about recognizing the signals produced by our nervous system.

That’s how the idea for Black Swan was born. A startup where we apply brain-computer interfaces to control equipment in combat conditions. First and foremost, we’re talking about FPV drones (First Person View drones equipped with a camera that shows everything from the pilot’s perspective – note by Digital Business).

– You moved away from aesthetic medicine because it didn’t align with your values. But isn’t the defense industry an even more ethically ambiguous space?

– I made a very deliberate choice to move into DefenseTech. These are technologies designed for defense and protection. Their purpose is to reduce the risk of human error, improve safety, and make systems more resilient, not to fuel aggression.

I used to try to stay away from politics and anything connected to it. But at some point, I realized that was a childish position. It’s easier to close your eyes and pretend problems don’t exist. What really matters is understanding what conflicts are happening in the world and why. There’s no such thing as neutrality here. I made my choice. In March, we started actively working with experts and assembling the first version of the product.

The Black Swan team put together a working neurointerface prototype in about a month. We used existing BCI sensors (devices that read the brain’s electrical activity from the surface of the head – note by Digital Business). Based on those, we built our own neuro-helmet that could be connected directly to drones. Then came the hardest part. Training the system to interpret brain signals and a person’s mental state in a way that could be translated into clear commands for the hardware. It wasn’t easy. Hardware projects are always more complicated than software ones. A lot didn’t work on the first try, so we had to keep tweaking and reconfiguring the equipment. At one point, we even crashed a couple of drones in the process.

«I wouldn’t recommend this path to anyone»

– What are the limitations of traditional drone controls, and how does Black Swan help pilots overcome them?

– The operator sees the image as if they were inside the aircraft. When a drone pilot is fully focused on the screen, it’s easy to make a critical mistake. The problem becomes even more pronounced when VR goggles are involved. At that point, the pilot can no longer see their own hands or the controls. It’s easy to mix up commands, press the wrong button, lose orientation, drop the controller, or even lose balance and fall.

In the military, this is especially critical. According to various studies, a significant share of service members who have seen combat experience PTSD (post-traumatic stress disorder – note by Digital Business). A drone pilot operates under constant pressure, making split-second decisions, and any mistake can cost lives. Not just their own, but others’ as well.

It’s much safer and more precise to shift control either to brain signals or to sensory and gesture-based interfaces. That’s exactly what we provide with our neuro-helmet and specialized wristbands.

— What makes your devices unique?

– Non-invasive neurotechnologies have actually been around for more than a decade. They usually look like lightweight neuro-helmets, headbands, or sensor rings worn on the head that read the brain’s electrical activity through the skin. They’re widely used in medicine. For treating depression, anxiety disorders, sleep issues, and for post-injury rehabilitation. Another major area is sports and eSports. There, brain-computer interfaces are used to train focus, reaction time, and performance under pressure.

The problem was that, until recently, these solutions worked reliably mostly in lab conditions. As soon as a person started moving, entered a noisy environment, or experienced stress, the signal became distorted. That’s exactly why many well-known startups, like Neuralink and Synchron, moved toward invasive solutions, implants that are surgically placed directly into the brain.

Instead of trying to «read a pure thought», we focused on recognizing intent. We combine brain signals with body movements and a person’s overall state, and AI brings all of that together into a single control model. That’s what allows the neurointerface to work reliably outside of lab conditions.

Black Swan is a hardware startup, but we’re not reinventing the neuro-helmet from scratch. We work with existing devices, assembling and fine-tuning the hardware for real-world scenarios. The core of the product is the AI model. That’s the piece that brings together neural signals, body movements, and a person’s state into a single control system.

– What does this look like in practice?

— Like real-life telekinesis. But there’s no magic involved, just science. When a person is fully focused, high-frequency brain waves become dominant. The system picks that up, and the drone starts moving. As soon as attention drops, the system detects it, and the drone stops.

Hand movements are used as an additional layer of control. Sensors on the wrists pick up simple, intuitive actions: a turn of the hand, direction of movement, clenching or unclenching the palm.

In about 95% of cases, the drone responds exactly the way the operator intended. It stops, keeps moving, or executes the command correctly. In the remaining 5% of cases, the command doesn’t go through. The system detects insufficient focus, a spike in stress, or «noise» in the signal and switches the drone into a safe mode. Most often, that means stopping or holding position. To reduce those cases, we’re continuing to improve the accuracy of the AI model.

– Do you already have early users?

– We have five clients who are testing the technology in pilot mode. For example, one woman is buying the device to organize robot battles. Others are using it to control drone-based filming at events.

We also regularly host drone races in San Francisco. Participants take turns putting on the neuro-helmet and controlling a drone purely through concentration. As long as a person stays focused, the drone keeps moving. The moment their attention slips, it stops.

At our most recent event, we had 30 participants and generated around $5,000 in revenue. What’s important is that we don’t handle the organization ourselves. We have a strong partner who takes care of all the operational side of things.

— How do entertainment competitions fit together with the defense sector?

– Events give us data that we use to train the model. Essentially, every race becomes a training ground for the technology. Right now, we have around a thousand neurodata recordings from different people. They help us understand how to further develop and scale the technology.

At the same time, the project’s core focus remains DefenseTech and working with defense and government clients.

– Defense ministries in any country tend to be very closed-off institutions. How do you actually manage to work with them?

— To be honest, it’s almost impossible. In the US, government defense organizations rarely work with foreign founders. There are really only two realistic scenarios. Either you have a truly unique, mission-critical technology that no one else can build, or you have people inside the system who are trusted and willing to vouch for you.

We have a former military officer on the team. He’s currently based in Washington, DC and helps us build relationships with key US defense institutions. Even with that, the level of distrust is still very high. That’s why the main challenge isn’t just demonstrating the technology, but building trust and long-term relationships. It’s a slow process that takes time, reputation, and the right people around you.

I expect it will take at least two to three years before we land our first contract.

I wouldn’t recommend this path to anyone. It’s difficult, risky, and requires an enormous amount of inner resilience.

– What can you offer these organizations that’s truly unique?

– Our edge is the data and the way we train the system to interpret human behavior in real-world, noisy, high-stress scenarios. That applies both to controlling different kinds of devices and to operating complex equipment. For example, certain types of helicopters, as well as monitoring a user’s vital signs like blood pressure and heart rate while they’re using our device.

«We’re planning to close a $2 million pre-seed round in 2026»

— If your long-term goal is to work with large government and defense organizations, what business model are you aiming for? Is it one major contract that covers everything financially, or multiple streams?

— Our main expectations are indeed tied to defense-sector contracts. In our case, we’re primarily talking about the US Department of Defense and its research agency, DARPA. For them, amounts like $100,000 to $200,000 for research are essentially operational expenses. Almost pocket change.

For us, the ideal scenario is a strategic partnership with the US government. That would open the door to major contracts. Not hundreds of thousands anymore, but budgets in the millions and even billions.

As of today, Black Swan has several letters of intent. Under their terms, once the pilot projects are completed, partners are considering buying out the technology and ordering around 100 devices.

— What’s the project currently running on, and what does the investment situation look like?

— For a while, the project was developed using the Friends, Family & Fools model. That means personal funds, plus support from friends, family, and people close to us who believed in the idea at a very early stage. We tried to keep costs as low as possible.

For example, I won a spot in one startup program, beating out around a thousand applicants from all over the world. That gave me access to free housing, a workspace, and a lab.I was also accepted into AlchemistX, one of the world’s leading B2B accelerators, as well as the Silicon Valley Residency program, supported by Astana Hub and hosted at Silkroad Innovation Hub in Silicon Valley. The program made it possible to connect with other founders and experts, and to get a much better understanding of the US startup ecosystem and market expectations.

We’re covering part of the costs through events and our first sales. We haven’t raised venture funding yet. That would definitely be easier if I were American. There is interest in the project, and it’s been consistent. DefenseTech remains a highly relevant field and will most likely become even more in demand in 2026.

Right now, I’m in talks with an angel investor who has a fairly large check size and strong connections within the industry. When it comes to the pre-seed round, our target is around $2 million. We’re planning to close it in 2026.

— What other plans does the startup have?

— We plan to keep hosting drone races every month. For me personally, it’s also a way to stay in a strict rhythm. With my ADHD (attention deficit hyperactivity disorder – note by Digital Business), deadlines work better than anything else.

Each event is a test, a source of data, and a chance to improve the product. At the next race, we’re planning to add a second participant. Right now, only one drone can be in the air at a time. We also want to make the course more challenging and introduce new levels, like flying over obstacles.

In the long run, the goal is contracts with the US Department of Defense and possibly a shift toward selling anonymized data. Data that, as they say, is the new oil. Essentially, this is a new type of insight into human behavior and brain activity under stress. The measurements we’re collecting now will become one of Black Swan’s most valuable assets in the years ahead.