nikolay donets

REvolut

The Research and Applied AI Summit (RAAIS) is a community for entrepreneurs and researchers who accelerate the science and applications of AI technology. In the run up to our 10th annual event on June 12th 2026 in London, we’re running a series of speaker profiles to shed more light on what you can expect to learn on the day!

At RAAIS we have a focus on translating cutting edge technology and research into production-grade products for real-world problems.

Nikolay runs Machine Learning Engineering at Revolut, where his organisation builds the AI platform behind the company’s production models, from classical ML to GenAI. His teams work on the governance, evaluation, release, and cost controls needed to keep AI running in production across a regulated financial product environment. Publicly, he has described this work in terms of scaling AI beyond a centralised model, giving teams shared platforms and tooling to ship AI systems more effectively across the company.

Building one platform for builders, operators, researchers, and compliance

A central theme in Nikolay’s work is that the hard problem in production AI is not building a model in isolation. It is building one platform that can serve builders, operators, researchers, and compliance at the same time. That framing is especially relevant now because many organisations have already learned that strong model performance does not by itself solve deployment. The harder challenge is creating infrastructure that supports evaluation, release discipline, governance, and cost control without slowing iteration to a halt.

For an ML audience, this is where a great deal of the field’s practical difficulty now sits. Production AI in regulated settings requires systems that can be monitored, governed, and updated repeatedly under operational constraints. In that sense, the platform is not separate from the model work. It is what determines whether model work becomes durable capability inside a real organisation. That is exactly the layer Nikolay’s role speaks to.

Why this matters now

A lot of the hardest work in AI has moved out of the model itself and into the systems around it. As organisations deploy more models across more products, the bottleneck becomes governance, evaluation, release discipline, monitoring, and cost control. That is especially true in regulated settings, where production AI has to satisfy not only technical teams, but operational and compliance requirements as well.

This is why platform work matters. It is the layer that determines whether model progress can become repeatable, scalable, and usable in the real world. As more companies try to support both classical ML and generative AI in production, those questions are becoming central to how AI is actually adopted.

Nikolay’s background

Nikolay holds a PhD in engineering, and his earlier work focused on control systems and predictive maintenance for critical infrastructure, giving him a background in engineering problems where reliability, monitoring, and operational discipline matter.