Nikolai Zhivotenko

nikolai@zhivotenko.com

Moscow, Russia

I’m a Principal AI/ML Engineer and systems architect working at the intersection of artificial intelligence, robotics, and large-scale engineering platforms.

My career began in solid-state physics and computational materials science, where I developed deep expertise in multiphysics modeling, composite materials, and numerical simulation. Over time, this foundation evolved into building AI-driven systems that operate not only in software environments, but in physical, real-world systems.

Today, I specialize in designing scalable AI architectures that bridge digital and physical domains — robotics, industrial platforms, embodied intelligence, digital materials sicence and production-grade ML systems. I translate complex physical and mathematical models into reliable, deployable intelligence.

I approach AI not as isolated software, but as intelligence embedded in physical systems. My background in physics and engineering enables me to build architectures grounded in real-world constraints — systems capable of operating in robotics, manufacturing, and large-scale industrial environments.

I’m driven by continuous learning and by building intelligent systems that connect the digital and physical worlds.

⚡️ breaking news

🪸	Jan 9, 2025	Thrasos 3D turns seaweed into 3D-printed reefs in Mexico!
🇷🇺	Feb 1, 2024	I relocated to Russia. Stay tuned for blog updates.
🏗	Aug 16, 2023	Mighty Buildings - Way to Fix The Future
📃	Nov 17, 2022	Added one of the earliest projects with my participation
🎓	Oct 15, 2022	Successfully completed a course at the University of Michigan
🇹🇷	Oct 1, 2022	I relocated to Turkey. Stay tuned for blog updates.

selected publications

Ritz method

N. Zhivotenko

Oct 2021

Abs HTML PDF

Calculus of variations is a mathematical discipline devoted to finding extreme (largest and smallest) values of functionals – variables that depend on the choice of one or more functions. Calculus of variations is a natural development of the chapter of mathematical analysis devoted to the problem of finding the extremums of functions. The emergence and development of the calculus of variations is closely related to the problems of mechanics, physics and other technical sciences. Methods of calculus of variations are widely used in various fields of mathematics. For example, in differential geometry, geodesic lines and minimal surfaces are searched with their help. In physics, the variational method is one of the most powerful tools for obtaining equations of motion, both for discrete and distributed systems, including for physical fields. The methods of calculus of variations are also applicable in statics. There are two main types of methods for solving variational problems. The first type includes methods that reduce the original problem to solving differential equations. The alternative is the so-called direct methods. These methods somehow solve the original problem of finding a function in a given class that would deliver an extreme value to a given functional. One of the most popular methods of this class is the Ritz method.
Numerical modeling deflected mode in the typical construction of small-size spacecrafts from the composite material with rational reinforcement

N. Zhivotenko, E. Eremina, and A. Sergeeva

Innovative Development Dec 2017

Abs HTML PDF

Rationally reinforced products from composite materials are of huge interest to application designing of small-size spacecrafts. Results of numerical modeling of the deflected mode of a standard construction of the small-size spacecraft are presented in this work. It is represented that the weight efficiency of rationally reinforced composites is three times higher than at traditional aluminum alloys.
Crossover from an ordinary ferroelectric to a relaxor ferroelectric in Sr₍₂₊ₓ₎Bi₍₄₋ₓ₎Ti₍₅₋ₓ₎NbₓO₁₈

S. Gridnev, N. Tolstykh, N. Zhivotenko, and 1 more author

Bulletin of the Russian Academy of Sciences Physics Oct 2016

Abs HTML PDF

Sr₍₂₊ₓ₎Bi₍₄₋ₓ₎Ti₍₅₋ₓ₎NbₓO₁₈ materials with x = 0.2, 0.4, 0.6, 0.8, and 1 are synthesized. Investigations of the dielectric constant of the samples reveal an increase in the blurring of the phase transition as the share of titanium ions substituted for niobium ions grows. The Burns temperature and the temperature corresponding to the dielectric permittivity maximum are determined for all of the above compositions. Crossover from a normal ferroelectric to a relaxor ferroelectric in Sr₍₂₊ₓ₎Bi₍₄₋ₓ₎Ti₍₅₋ₓ₎NbₓO₁₈ samples is observed.