Ernest is an experienced Software Engineer with 7+ years of strong technical competencies and hands-on expertise in the development and delivery of mission-critical Java solutions and services on projects for leading tech companies in Thailand and Russia. He has team experience, is a good communicator and uses a business-oriented approach for interactions with clients.
Efficiently led a team of five developers and data scientists to develop and rewrite a fraud detection service that handles bookings (100,000+ per day) through the system.
Oversaw the migration of the fraud detection system from a manually-created set of heuristics to new machine learning models.
Implemented an ML feedback loop using historical data storage and newly-written tools and increased performance by 35%.
Supervised the replacement of legacy code and outdated technologies on the real-time module of the Crypta system (user behavior analysis for ad targeting) for Yandex.
Utilized new approaches and tools to engineer A/B testing of the geotracking algorithms on the system.
Created a model that predicts users’ income based on mobile phone activity and introduced various cross-device identification methods to the system for the client.
Designed the project architecture and implemented an initial version of the context advertisement automation service on the project.
Provided guidance and technical expertise for the development and enhancement of dozens of different microservices that process and supply data to a search engine optimization system for the client.
Successfully implemented various social network crawlers and rewrote the user payment module.
Collective is the creator of the largest global network of enterprise buying behaviour using data, artificial intelligence and predictive technologies to guide sales professionals through the activities that lead directly to revenue. Co-designed and implemented the back-end of the module. The service provides recommendations and allows users to connect (send, manage, and reply to requests) with each other. The clients of the service are various front-end pages.
Crypta is Yandex's own proprietary behaviour analytics technology which enables advertisers to target ads to a specific audience. This technology allows for the classification of web users based on online behaviour. Added support for A/B testing to the real-time system processing about 200,000 PRS. The location is used for ad targeting, which is a core business for the company, making the detection an essential part of the system.
Developed a Java annotation-driven API client on the solution. The annotations allow the client to be extremely readable and hides implementation details.
Education
MSc. Computer Science
Politecnico di Milano
2018 - 2020 (2 years)
BSc. Software Engineering
Vysshaya Shkola Ekonomiki / Higher School of Economics
2011 - 2015 (4 years)
Associate Oracle Certified Professional, Java SE 7 Programmer
Oracle
2013
Oracle Certified Professional, Java SE 7 Programmer