Sergey Melnik, Phd Email & Phone Number

Toronto, Ontario, Canada

Toronto, Ontario, Canada

Lead a team of Data Scientists and Data Engineers who develop new services to solve business problems.

Canada

Toronto, Canada Area

- Responsible for Data Science and the corresponding parts of cloud architecture- Work closely with internal clients across the company to identify and solve business challenges- Generate, lead, execute Data Science projects to obtain actionable insights or create new/improved products- Act as a leader/mentor or in hands-on capacity depending on the need and complexitySIGNIFICANT PROJECTS:- Led the design of data-processing pipelines for a near-real-time data analytics project. Developed several POCs using different AWS technologies, solved multiple technical design challenges that resulted in performance improvements (up to 275 times faster than initial designs) and significant savings on AWS cloud infrastructure- Developed NLP algorithms for the text analysis of customer complaints: sentiment analysis (with LSTM neural networks), complaint classification, top issues discovery, tracking of any complaint category- Created a Sales Analysis Framework for Events based on signal decomposition techniques tosystematically analyse the impact of various events on sales and identify cause and effect relationships- Developed sales prediction models (time series forecasting) for improved labour scheduling- Came up with new Customer Churn Indicators and Lifetime Value calculations for CRM. Reduced I/O and compute time from days to less than 1 hour, thereby making possible daily updates of the CRM system

Toronto

- Developed recommendation algorithms.- Came up with novel designs and implemented algorithms in Apache Spark to achieve 4-34 times speed-ups over implementations based on the standard Spark Machine Learning library, while solving memory issues.- Created a Deep Learning classifier in Keras/Tensorflow, trained on GPUs. Outperformed “classical” NLP by 15%.

Ireland

- Conducted research, developed novel algorithms, models and analytical methods- Supervised graduate and undergraduate students- Led and participated in problem-solving workshops with industrial partners- Published in top-level journals and presented at international conferences- Developed international collaborations- Taught maths to students- Main areas of expertise: social influence and adoption cascades on networks (e.g., thenetwork of Facebook friendships), systemic risk in financial systems, data analysis.

Oxford, United Kingdom

- Developed a novel model for multi-stage complex contagion- Demonstrated new dynamical behavior, which cannot occur in single-stage models- Developed a new ensemble of modular networks (with heterogeneous correlations) andan analytical approach to modelling dynamics on such networks- Published the results in international peer-reviewed journals and presented at conferences

Ireland

- Modelled dynamics on clustered networks- Analysed the effect of clustering on critical network dynamics- Discovered high effectiveness of tree-based theories for networks with clustering- Results published in international peer-reviewed journals and presented at conferences

Samara Region, Russian Federation

- Developed software for data analysis and control of the device measuring the gascontents in metals for AvtoVAZ (Lada) car manufacturer- Implemented with Microsoft Visual Studio (C++) and MS Access database via ODBC

Complex Systems Summer School

Doctor Of Philosophy (Phd), Physics

Activities and Societies: Tae Kwon Do Club, Karate Club, SwimmingPhD Thesis: “Nonlinear and Stochastic Dynamics of Semiconductor Laser.

Bsc & Msc, Computer Science & Mathematics, 1St Honours (Score: 98.8%)

Thesis title: “Computer simulation of emergency situations on oil plant facility 43-102” My BSc and MSc were taken as a single 5-year.