Ruslan Mozhaev

AI for healthy longevity — from population patterns to the individual.

Independent researcher. I map what the evidence on cognitive and healthy ageing actually supports, and work towards AI that can carry it from population averages to a single person. The AI side goes back to my 2009 PhD research on multi-agent self-learning systems.

RESEARCH / what I work on

01

An evidence map of cognitive longevity

Fourteen interacting systems — sleep, vascular health, learning, social ties and more — with every link graded by the strength of its evidence. The map keeps apart three things that are usually blurred: no data, no effect at adequate power, and an effect only under certain conditions.

02

From population to the individual

A group average can show no effect while some individuals respond strongly, and a group effect can rest on a few strong responders. Such processes are generally non-ergodic: what holds for a population does not transfer to one person by default, so the bridge between the two levels has to be modelled rather than assumed. This is the question the AI work is built around.

03

Multi-agent, continual learning

Architectures that keep learning from one person without forgetting what they know about everyone else — the engineering side of that bridge.

BACKGROUND / the long way here

The thread starts in 2009 with PhD research on multi-agent self-learning systems and runs alongside twenty years of building production software.

2006→ Twenty years of production software — enterprise and mobile systems, lately AI integration.
2007 Master of Engineering in Computer Systems, Riga Technical University — specialising in artificial intelligence and system analysis.
2009–10 PhD research on holistic multi-agent self-learning systems under Prof. Jānis Grundspenkis.
today An evidence map of cognitive longevity, and the population → individual question that the AI work will build on.