How an Uzbek is building the future of humanoid robots and attracting millions in Silicon Valley

Sanzhar Atamuradov is creating one of the most ambitious platforms in global robotics — Humanola, a Silicon Valley startup that turns telepresence into the foundation of physical artificial intelligence. The company allows people from all over the world to control humanoid robots with minimal latency, and every operator action is transformed into training data for future physical AGI.

In a joint column by IT Park Uzbekistan and The Tech, Sanzhar spoke about how his journey unfolded — from a boy herding sheep under the stars of Samarkand to the founder of a Silicon Valley startup that raised $2.5 million to build the foundation for global robotics.

Sanzhar Atamurodov, Samarkand city, founder of the startup Humanola, LinkedIn

About myself

I was born in Samarkand — a city where childhood passes under endless stars and where dreams, for some reason, feel closer. From a very early age, I was drawn to technology: I took apart everything I could take apart, tried to understand how mechanisms worked, and later — computers and robots.

After school, I went to South Korea to study at KAIST — a leading university in engineering and robotics. There I studied computer science with a focus on robots, created autonomous systems and drones. This experience became a starting point for me: I realized that I want to work with humanoid robots and artificial intelligence for my entire life.

During my studies at KAIST, I tried to focus more on AI and robotics. During academic semesters, I spent most of my time studying the theoretical aspects of AI and robotics, and during breaks I went to companies for internships. In Korea, there are two vacation periods per year: summer and winter — each lasting about two and a half months. This gave me a lot of time to work on interesting projects in companies.

First steps in the profession: computer vision and autonomous robots

After returning to South Korea, I interned at StoneLab Inc, where I participated in developing a medical diagnostic application based on computer vision: I performed data labeling, neural network training, and integration of the model into the backend. At the same time, I created an API for biometric authentication — my first system used by real users.

Work at Macroact Inc. became the next important stage. There I worked on autonomous navigation for a quadruped robot, algorithm optimization, behavior tuning, and setting up Gazebo simulations — a program for creating and running virtual worlds with robots and objects, allowing testing without risk to real equipment. This became a huge springboard for moving to more complex robots.

During my studies at KAIST, I also tried myself as a researcher at AVE Lab and IRiS Lab at the university. At AVE Lab, I worked on algorithms for autonomous vehicles. At IRiS Lab, I worked on perception systems for autonomous transport — from processing LiDAR data to lane detection. This stage gave me incredible depth of understanding of how a robot “sees” the world.

Industry: from warehouse AMRs to quadruped robots

After research, I moved into industry. At Digitrack Inc, I worked on autonomous mobile robots for warehouse automation.

Later, at Raion Robotics, I worked on locomotion of quadruped robots. It was a real challenge — to make the robot move stably, adapt to surfaces, and perform complex maneuvers.

Studies and work in the USA: integration of engineering and science

I moved to Atlanta and enrolled at the Georgia Institute of Technology — one of the strongest centers in the world for robotics and artificial intelligence.

Studying in the USA became a turning point for me. Here I gained deep fundamental knowledge in robot control, neural networks, perception systems, and autonomous behavior. I took courses on robotic manipulators, motion optimization, machine learning, real-time algorithms — everything that forms the foundation of modern robotics.

At the same time, I worked as a robotics engineer at Atlanta Ventures, where I created mobile robots for security and inspection tasks. That’s when I began to feel how theory connects with practice — how knowledge from classrooms instantly turns into code, algorithms, and moving machines.

After that, I continued my research career at Georgia Tech, working on humanoid loco-manipulation — the ability of humanoid robots to coordinate movement: walking, lifting objects, opening doors, performing complex sequential tasks simultaneously.

And it was at that moment that it struck me: large language models made a breakthrough thanks to massive amounts of data from the internet. But robots do not have such an “internet” — they are literally starving for data. I realized that the only scalable way to create high-quality training data is telepresence, where an operator controls a robot. But no one in the world was doing this systematically.

That’s when I had the idea that changed my life: to build a platform that would give every roboticist a telepresence tool and create data infrastructure for physical AI. That’s how Humanola was born in June 2025.

We are still in the early stages: mass adoption for physical labor automation has not yet arrived. Humanoid robots working in factories, warehouses, or homes in the USA do not yet exist — but this is about to happen very soon.

What problem Humanola solves

Today, data for training robots is isolated within companies and laboratories. Everyone creates their own small datasets, no one shares, and the industry is literally stalling. It’s like a world where every AI solution of 2023 would be trained on its own small local internet.

Before Humanola, every company built its own data collection infrastructure — expensive, slow, and inefficient. Many simply worked with limited datasets, which slowed development. There was no unified ecosystem that combines data and provides access to it.

Our team is inspired by a future where robots perform dangerous, routine, and monotonous work, and humans focus on creativity and strategy. The only barrier between today and that future is the lack of data for physical AI. By solving this problem, we accelerate the future by years.

I realized that the problem is not only fragmentation of individual solutions, but the absence of an industry platform that all market players could connect to. Duplication of efforts and data isolation slow down the entire field. This understanding became the driving force behind the creation of Humanola.

What Humanola does

Humanola develops a platform for remote robot control and a comprehensive infrastructure for collecting, processing, and analyzing physical data.

The platform consists of two key components:

— operators in VR headsets control robots in real time with minimal latency

— all session data is automatically collected, cleaned, labeled, and turned into ready datasets for AI training

As a result, companies get a powerful tool to accelerate the development of their robots.

Humanola is the only independent platform not tied to specific hardware and providing a full cycle: from robot control to full data processing. Other solutions require building your own infrastructure or do not provide this level of integration.

Among the clients are companies implementing robots in logistics, agriculture, and warehousing, as well as developers and researchers who need fast access to high-quality data.

On challenges

The most difficult part was achieving minimal control latency over long distances and integrating different types of hardware. This required deep technical improvements at all levels — from the VR interface to network protocols and cloud infrastructure.

We solved this using our own real-time network infrastructure based on UDP protocols with error correction and adaptive streaming, similar to top VR games. For long-distance control, we deployed regional edge servers to reduce latency to within 80-100 ms even between continents.

Hardware integration was ensured through a unified API for robots — we abstracted joint control, sensor streams, and safety protocols so that any humanoid or manipulator could connect with minimal setup.

The breakthrough happened when we combined private 5G networks in test environments with GPU-accelerated video encoding/decoding on both the operator and robot sides. This enabled stable control with latency below 100 ms — for example, from Tashkent to San Francisco. Thus, we proved that global scalable telepresence is real.

The main motivation not to give up is belief in the mission and understanding that we are building not just a product, but the foundation for a new generation of AI and robotics.

A technical breakthrough in remote control with latency under 100 ms became critical — it was at that moment that the concept became reality and the product entered the market.

Achievements

What I am most proud of in our work:

1. Real clients who note a radical improvement in their work. To date, we have two B2B clients — both companies develop humanoid robots and actively use our platform for telepresence and data collection. We made two early-stage sales, which confirms the value of the solution: clients note a 30-40% acceleration in development due to ready-made datasets.
   
2. Rapid transition from idea to industrial platform. First, remote telepresence allows clients to instantly deploy robots and control them remotely from anywhere in the world. This means companies can start working with robots immediately, without a long period of developing local control systems. Second, our data processing pipelines handle massive volumes of data and transform raw sensor data into datasets ready for model training. This allows companies to achieve robot autonomy much faster than with traditional approaches.
   
3. Building a strong team around a big mission. Today, our team consists of five people: three co-founders and two engineers. I lead the company as CEO, managing both the technical side and the strategic vision. Our CTO is Zhaoyuan Gu, PhD, with whom I conducted joint research on humanoid robots at Georgia Tech. Our COO is Akbar Erkinov, responsible for operations and ensuring efficient functioning of the organization.

We completed a seed round of $2.5 million from American venture funds and angel investors, the first clients are already implementing the platform, and we see confirmation of the relevance of the problem and the value of the solution. Revenue and user numbers are not disclosed yet due to the early stage.

We are supported by leading venture funds, and our clients are companies seriously investing in robotics. Adoption dynamics are the best proof of effectiveness.

On development in the USA, Korea, and Uzbekistan

The USA is the leader in developing the “brain” of robotics — artificial intelligence and software systems that allow robots to think and act like humans. This is what drives the industry forward.

Korea has strong teams and constantly pushes the boundaries of technology. However, its scale is significantly smaller than in the USA. Korean companies often focus on hardware and innovation but lack the scale of the American market.

Uzbekistan is only at the beginning of its journey. Robotics here is at an early stage of development.

Plans

We aim to integrate Humanola into all major humanoid robot platforms in the USA and worldwide, and also to become an infrastructure partner for industry leaders — Nvidia, Tesla, Google — in building physical AGI.

As part of our expansion strategy for the first 12 months, we are targeting large-scale robot deployments through robotics manufacturers. We are already hiring telepresence operators from Uzbekistan who will remotely control robots operating in the USA and other countries.

This creates huge economic opportunities — we are talking about thousands of new jobs in regions of Uzbekistan, while simultaneously scaling the volume of collected data. Operators can work from anywhere in the world, robots can perform tasks anywhere in the world, and the data will return back for training increasingly advanced models. This is a win for the entire robotics industry, for job creation in emerging markets, and for the development of physical artificial intelligence.

We want to accelerate the development of physical AGI — to make robots mass assistants and permanently change the approach to labor. Our mission is to create infrastructure that democratizes the industry and makes intelligent robots widely accessible.

Advice to entrepreneurs

Dream big and do not be afraid to take risks. All significant achievements come on the edge of failure — the main thing is perseverance and consistency.