Probability theory, Computer methods for big data analysis

Results: - researched and tested a new RecSys model for personalized recommendations on the core product;- developed a new online recommender with serving and continuous learning platforms based on the Ray framework to bring the new model into production;- achieved significant business impact: +17% engagements from the recommendation feed, +7% product-wide interactions.

Project "Information resource about earthquakes, seismic impacts and hazards" (eqalert.ru)Project team: 5 developers.Scientific results: - Possible Connection Between Recent Seismicity and Fluid Injection in the Offshore Oil and Gas Field Area of Sakhalin Island, Russia (https://doi.org/10.1007/s00024-022-03006-y, read online: https://rdcu.be/cJ2O2)- Sakhalin real-time shaking map tool for industrial and civil facilities safety (read online: https://doi.org/10.1051/matecconf/201926503002)- Testing of a New GMPE Model in Probabilistic Seismic Hazard Analysis for the Sakhalin Region (https://doi.org/10.3103/S0747923919030095)- Regional Ground Motion Prediction Equation for Sakhalin Island (https://doi.org/10.3103/S0747923919010080)Tech results:- designed the micro service architecture of eqalert.ru;- developed database structure for manage earthquake-related data;- deployed docker-based stateless services for performing geospatial data calculations;- designed the seismic processing system (seisan, earthworm);- developed API, web-fronted, iOS/android sub-projects.Technology stack: vue.js, laravel/nova, mariadb, linux, kvm, vmware, nginx, php-fpm, firebase, telegram, js/php/julia/python/java

I worked on a standalone machine learning project in seismology. Publication: Stepnov, A.; Chernykh, V.; Konovalov, A. The Seismo-Performer: A Novel Machine Learning Approach for General and Efficient Seismic Phase Recognition from Local Earthquakes in Real Time. Sensors 2021, 21, 6290. https://doi.org/10.3390/s21186290 (read online: https://www.mdpi.com/1424-8220/21/18/6290)

I have worked as a statistician in various research projects:- regressed parameters and developed a probabilistic model to assess the risk of landslides caused by earthquakes (Article text: https://www.mdpi.com/2076-3263/9/7/305);- discovered missed events by analyzing waveform time series data using cross-correlation methods;- developed a probability model for predicting landslides based on precipitation characteristics (Read article online: https://rdcu.be/cpW0j);- designed experiments to test hypotheses about increased seismicity after the start of oil and gas productionPlease see my scientific publications for the results.Methods: linear/logistic regression, parameters estimation, probabilistic analysis, feature engineering, explanatory data analysis, cross-correlation analysis and hypothesis testing

Co-leader of the project "Development of new technologies for monitoring and managing seismic risks (natural and induced) in the industrial development of offshore oil and gas fields", Federal Special-Purpose Program of the Russian Ministry of Education. Project ID 14.607.21.0105 (RFMEF160714X0105). Project implementation period: 2014 – 2016. Number of participants: 15 developers/scientists. Grant amount: 15 000 000 RUB (~274 000 USD).Project results:The automated seismic monitoring system was implemented. Implemented such data processes as collection, transmission, storage and processing in real time. The system has been tested in North Sakhalin at a high level of industrial noise factors. The system preforms earthquake location within ML≥3.0 without missing events and detects seismic signals in data flow for events with МL≥1.0. The one of disadvantage of the system is high rate of false triggered events that increases as number of seismic stations grows.Technology stack: seisan, earthworm, winston/swarm, obspy, linux, kvm, zabbix, ocfs2, nfs, C/bash/python, mysql.

Leader of the project "Study of the seismic regime of the Sakhalin active fault zones during the industrial development of oil and gas fields on the shelf." Russian Foundation for Basic Research (RFBR). Project ID 14-05-31323 (mol_а). Project implementation period: 2014 – 2015. Number of participants: 2 scientists. Grant amount: 800 000 RUB (~14 500 USD).Project results:- detected seismically active zones using clustering methods;- justified completeness of the seismic catalog on the magnitude of earthquakes;- the errors has been estimated in determining the parameters of earthquake sources (lat, lon, depth);- defended my PhD Degree (one of the grant task): https://youtu.be/dNxSln9nttI;- implemented some routines for statistical analysis of earthquake catalogues.Methods: linear regression, parameters estimation, unsupervised machine learning

Leader of the project "Development of new software solutions and methods for assessing the dynamic parameters of the site response to ground shaking according to field geophysical and seismological observations. Innovation Promotion Fund (http://fasie.ru). Project ID 1842ГС1/26934. Project implementation period: 2014 – 2016. Number of participants: 7 developers. Grant amount: 2 000 000 RUB (~30 000 USD)."Project results:- The GMPE model for Sakhalin has been developed and well tested. That project is available under the Apache license https://github.com/geophystech/GroundMotion.jlMethods: linear regression, parameters estimation, data acquisition

Tech leader of the project "arsenalpay.ru" (e-commerce service provider). Project team: 4 developers, 1 engineer. Results:- more than 30 virtual machines designed and deployed to service project tasks (Linux/XEN, PostgreSQL, nginx, php-fpm, bind, snort, Zabbix, NFS, FreeRADIUS);- deployed L3 VPN services (Cisco webvpn, IPSEC);- implemented security tasks for the entire IT infrastructure of the company for PCI DSS certification;- supervised the development of API (PL/pgSQL), web-frontend (PHP/Yii), client side service integrations.

Tech leader of the project "IPTV service" (~10K active subscribers). Project team: 2 engineers. Results: - built a headend for receiving satellite TV/radio broadcasting and streaming it on IP multicast (MPEG2/MPEG4, special hardware by Motorola);- carried out Integration with local TV and radio companies (L2/L3 networking based on Cisco equipment).In this company, I also supported the SMSc service with ~500K active subscribers (SUN/Solaris, Oracle Database).

The project "Highload web services for DSL backbone" (additional services for Internet users). 100K+ users/day.Project team: 4 developers, 1 engineer.Results:- designed and launched simple fault-tolerant web-cluster using php-fpm, nginx, and postgresql;- implemented svn/redmine in the development process;- launched a video hosting service, a file sharing service, a news site.The stack of technologies: drupal, bash, nginx, php-fpm, postgresql, postfix, zabbix, iptables, svn/redmine, ocfs2, linux, xen, freebsd.

Tech leader of the project "Information and telecommunication resources of the Far East" (Far Eastern Branch of the Russian Academy of Sciences). Project team: 5 engineers. Experience:- built network backbone (L2/L3) for Sakhalin branch (Cisco equipment);- deployed core services for Sakhalin branch (Linux/XEN, FreeBSD, postfix/Cyrus-IMAP/amavis/ClamAV/spamassassin, squid, ProFTPd, Samba, mrtg/zabbix).

The project "Digital Language Center". Project team: 2 engineers, 3 assistants.Results:- built L2 network segment for new digital laboratories;- a simple file-based database was developed for the digitization of Soviet educational materials;- developed standards for working with digital laboratories.The stack of technologies: cisco L2, windows based workstations, linux, NFS, Samba.

Project name: "The Univercity LAN".Project team: 2 engineers.Results:- launched the workstation management system (100+ PC's) based on MS Windows 2000 Server;- launched the University firewall based on MS ISA server.The stack of technologies: cisco L2, MS Windows 2000, MS ISA server 2004/2006