Looking for a job

Roman Okhotnikov

@romanokhotnikov

Message

I'm building machine learning systems for mobile networks, network reliability, and time-series prediction.

Russian Federation

Message

What I'm looking for

I'm looking to build practical machine learning systems for complex real-world data, especially network reliability, forecasting, anomaly detection, and graph-based analysis.

I'm building machine learning systems at Beeline for mobile-network KPI analysis, anomaly detection, traffic forecasting, and packet-loss investigation. Using base-station and transport-network adjacency data, I help identify underlying causes and pinpoint problematic network hops.

Previously, I supported Beeline's IP Backbone research, migrating IP address management from Oracle to NetBox, parsing router configurations, and analyzing crash data. I used graph neural networks to identify weak infrastructure points, model crash propagation, and estimate device-crash probabilities up to a week ahead.

I've also worked on trading-bot tooling and Bayesian optimization at AIM Fund, improving backtesting time and results by 33%. My projects include LLM-based recommender-system benchmarking, time-series forecasting with RNNs and GNNs, and EEG/MEG classification using classical ML features.

Experience

Work history, roles, and key accomplishments

Current

Radio Access Network ML Expert

Current

Beeline

Aug 2025 - Present (1 year 1 month)

Conducted mobile network KPI analysis using base station adjacency data and implemented error prediction models with probabilistic models. Designed and implemented Airflow-orchestrated tasks for anomaly detection and KPI time series prediction, achieving reasonable accuracy on hourly data while preserving computational resources.

Python SQL Airflow Unsupervised Learning Time Series Analysis Anomaly Detection Network Analysis