Borislav Mavrin
NLP EngineerAI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Senior MLE @ Electronic Arts
- First to reproduce gpt-oss-20b SWE Verified results with native Harmony Agent https://arxiv.org/abs/2604.00362 https://github.com/borislavmavrin/harmonyagent - Building agents at Electronic Arts - Developed automated RAG evaluation framework that does not require labelled data and can be used on production client data directly (1 US and 1 Canadian patents pending). It is used internally and offered as a product in beta - Developed evaluation framework for testing LLM component of the production RAG system (clients: Dell, Zoom, Vanguard etc) which is used as a part of CI/CD and migrations - Was first in the company to successfully apply Automated Prompt Tuning to the production RAG, which reduced the budget for migration from 84 to 12 person days and increased client answer rate/accuracy by 5-20%. - Migrated commercial production RAG system to 3 LLMs - Prototyped search agent and developed evaluation framework for search agent that helped to shape search agent internal roadmap - 3 master degrees: MSc in Statistical Machine Learning (UofA), MA Economics, MSc Math Physics. - 8 publications in ICML, AAAI etc., 3 US patents, 1 Canadian patent
Université de Montréal
Deep Learning Summer School 2016, Deep Learning
January 1, 2016 – January 1, 2016
University of Alberta
Master's degree, Statistical Machine Learning
January 1, 2015 – January 1, 2017
Concordia University
Master's Degree, Mathematics
January 1, 2013 – January 1, 2015
Concordia University
Master's Degree, Economics
January 1, 2008 – January 1, 2010
Electronic Arts (EA)
Senior MLE
January 1, 2026 – Present
Edmonton, Alberta, Canada · Hybrid
Coveo
Senior Applied NLP Scientist
July 1, 2023 – January 1, 2026
Montreal, Quebec, Canada · Remote
flyx.ai
Lead Machine Learning Researcher
May 1, 2023 – June 1, 2023
LG Electronics
Senior AI Research Scientist | Team Lead
March 1, 2022 – February 1, 2023
LG Electronics
AI Researcher | Team Lead
May 1, 2020 – May 1, 2022
LG Electronics
AI Researcher
December 1, 2019 – May 1, 2020
Huawei Technologies
Machine Learning Researcher
May 1, 2018 – December 1, 2019
Greater Edmonton Metropolitan Area
News360
Machine Learning Developer
January 1, 2018 – May 1, 2018
Greater Vancouver Metropolitan Area · On-site
Janalta Interactive Inc.
Intermediate Developer – Machine Learning
October 1, 2017 – January 1, 2018
Greater Edmonton Metropolitan Area · On-site
DriveABLE Inc.
Data Scientist (Internship during graduate program)
March 1, 2017 – August 1, 2017
Greater Edmonton Metropolitan Area · Hybrid
MaxFleet Solutions Ltd.
Data Scientist (Internship during graduate program)
January 1, 2017 – September 1, 2017
Greater Edmonton Metropolitan Area · Hybrid
Cell phone data distribution discovery.
January 1, 2016 – Present
One typical example of the spatial sparse recovery problem is to infer spatial distribution of cellphone activities based on aggregate mobile traffic volumes observed at scattered base stations. Extensive evaluations based on a large dataset of phone call records and a demographic dataset from the city of Milan show that our approach significantly outperforms various state-of-the-art approaches, including Spatial Spline Regression.
Stocktwits.com sentiment analysis (over 47,000,000 tweets).
January 1, 2016 – Present
Stocktwits is the social network for stock traders, where users can explicitly share their sentiment about specific stocks. The project’s goal is to identify useful signals for stock price predictions from users’ sentiment. I was able to identify trading strategies that exceed sp500 return by over 200%.
Cultural Fit Analysis
The candidate's diverse experience across various companies (EA, Coveo, LG Electronics, Huawei, News360) and project types (NLP, recommender systems, reinforcement learning, financial modeling) suggests adaptability and a broad interest in different applications of machine learning. Their academic background and research publications indicate a strong drive for continuous learning and contribution to the field, which aligns well with a research-oriented or innovative culture. The explicit mention of collaboration with universities further supports a collaborative cultural fit.
Soft Skills & Operational Fit
The candidate's resume indicates strong leadership, problem-solving, and teamwork skills through their team lead roles and project descriptions. Their experience in leading collaborations and growing teams suggests good operational fit for managing projects and potentially mentoring junior team members. The detailed descriptions of project outcomes (e.g., improving user engagement, reducing memory footprint) demonstrate a results-oriented approach.