Senior Developer · Explainable AI · Kyiv, UA

I build resilient software
& explainable AI.

I design software and explainable AI that's built to be understood, verified, and trusted.

Sixteen years building high-load platforms in PHP, Python, Go & JavaScript — now focused on explainable AI for cybersecurity: detection systems that are data-driven, verifiable, and built to be understood.

Hire me →View research
Oleksandr Mostovyi logo
01 / About

Engineering that lasts, research that looks ahead.

I've spent sixteen years building for the web — from custom CMS engines and e-commerce platforms to real-time dating networks serving millions. I lead teams, architect for scale, and stay close to the code.

Today I split my time between senior product work at Killing Kittens and doctoral research in cybersecurity AI — fusing machine learning with symbolic and formal methods to detect intrusions in satellite communication systems.

2023 — Present
PhD Candidate · Computer Science
Institute of Cybernetics, NAS of Ukraine
Neuro-symbolic intrusion detection research.
2018 — Present
Senior Developer
Killing Kittens Ltd
New dating platform & social network built from scratch; real-time notifications, profile matching & verification.
2014 — 2018
Team Lead
Woact LLC
Led the dev team; high-load Laravel / Yii platforms with AngularJS integration.
2012 — 2014
Senior Web Developer
Empower Network LLC
Zend Framework 2 apps, RESTful APIs, payment processors & search.
2008 — 2012
Web Developer
WebProService · Cupid PLC · Frontex
E-commerce, custom CMS, Neo4J graph search & functional modules.
02 / Stack

A deep, polyglot toolkit — chosen for the problem, not the hype.

Languages
PythonPHPJavaGoJavaScript
AI & Machine Learning
Machine LearningDeep LearningLSTM / sequence modelsSymbolic reasoningRule-based systemsSMT / symbolic execution
Frameworks
DjangoFlaskLaravelSymfonyZendSpringEchoNode.jsVue.jsReact.js
Databases
PostgreSQLMySQLMongoDBRedisNeo4JAerospikeClickHouse
Messaging & Search
RabbitMQElasticMQAWS SQSElasticSearchSphinx
Infrastructure
DockerKubernetesMicroservicesWebSocketsRPCReal-time systemsLinuxHPC clusters
Caching
MemcachedMemcacheRedis
03 / Research
Oct 2023 — Present · PhD Candidate, Computer Science

Neurosymbolic intrusion detection for satellite systems.

At the V.M. Glushkov Institute of Cybernetics, NAS of Ukraine, my doctoral research fuses machine learning with symbolic and formal methods to secure satellite communication systems — detection that is data-driven, explainable, and formally checkable against a defined threat model.

Sixteen years of high-load software engineering, now turned toward applied cybersecurity, formal verification, and machine learning for security.

Neurosymbolic IDSFormal verificationLSTM modelsSymbolic executionSMT solvingThreat modellingHPC pipelines
Doctoral programme
Research topic
Neurosymbolic intrusion detection systems for satellite communication networks
Supervisor
Oleksandr Letychevskyi
Defence
Oct 2027
Languages
Ukrainian (native) · English (C1)
Award · 30 Jan 2026
ActInSpace 2026 — Ukraine, Kyiv · Sponsor Prize
International space-innovation challenge by CNES (France) × ESA · Project: "Neurosymbolic Intrusion Detection System for Satellite Systems".
Line 01
Satellite intrusion detection

Neurosymbolic IDS for satellite communication networks — LSTM-based sequence models fused with graph-based logical inference. A pipeline that is data-driven and explainable, with detection logic formally checkable against a defined threat model.

Line 02
Binary vulnerability analysis

Symbolic execution, SMT-based reasoning and distributed execution on HPC infrastructure — including control-flow-graph recovery for dynamically loaded code via symbolic library resolution, applied to malware analysis and software supply-chain security.

Current directions
Protocol-level analysis of satellite link & ground-segment traffic — CCSDS USLP, SPP & others
Hybrid neurosymbolic architectures: deep sequence models + rule-based reasoning
Formal verification of detection logic & adversarial robustness of learning-based IDS
Symbolic execution & SMT-based vulnerability analysis of ARM64 binaries
Distributed binary-analysis pipelines on HPC clusters

Selected publications & talks

Peer-reviewed02
Control flow graph recovery for dynamically loaded code via symbolic library resolution
Bulletin of NTU "KhPI" — System Analysis, Control & IT
2026 · DOI 10.20998/2079-0023.2026.01.12 · DOAJ, ICI · sole-authored
Source code ↗
Neurosymbolic approach for attack detection in satellite communication systems
Problems in Programming · presented at UkrPROG-2024
2024 · DOI 10.15407/pp2024.02-03.223 · open access · sole-authored
DOI ↗
Under review04
A quantized deep learning approach to intrusion detection in narrowband-IoT non-terrestrial networks
ICTERI 2026 · PhD Symposium
A knowledge-graph-based neurosymbolic approach to network intrusion detection
ARTIIS 2026 · with T. Biloborodova
Neural constraint-dependency graphs for automated security analysis of embedded firmware
DESSERT 2026 · IEEE
Symbolic digital twin for intrusion detection in satellite communication systems
iTest 2026
Conference talks03
Distributed SMT-based vulnerability analysis of ARM64 binaries on HPC clusters
ICT 2026 · Apr 2026
Digital twin for ARM64 symbolic state modeling & vulnerability detection
IT: Theory & Practice 2026 · Mar 2026
Neurosymbolic IDS via LSTM networks & graph-based logical inference
UkrPROG-2025 · Kyiv
Education
2022–24
M.Sc. Computer Software Engineering
American University Kyiv (powered by Arizona State University)
Summa Cum Laude
2011–12
M.Sc. System Engineering · Robotics
Kremenchuk University of Economics, IT & Management
Summa Cum Laude
2007–11
B.Sc. System Engineering · Robotics
Kremenchuk University of Economics, IT & Management
Summa Cum Laude
04 / Contact

Let's build something
worth shipping.

Open to senior engineering roles, contract work, and research collaboration. Use the form below or one of these:

LinkedIn ↗ORCID ↗Download CV ↓