Victor Gueorguiev
Senior Machine Learning Engineer
Sofia, Bulgaria | victor.gueorg@gmail.com | (+359) 88 979 3111
LinkedIn: linkedin.com/in/victor-gueorguiev | GitHub: github.com/BrutishGuy
Summary
I have diverse experience as a data scientist and ML engineer, utilizing my analytical, statistical, and programming skills to collect, analyze, and interpret large data sets, and to build end-to-end AI solutions.
I have deployed actionable data-driven solutions, implemented impactful AI features, and solved multiple difficult business challenges. I hold a master's degree in AI and computer science. My wide range of technical competencies include statistics, deep learning, Bayesian optimization, reinforcement learning, NLP & LLMs, software engineering best practices, and cloud infrastructure & Kubernetes.
Work Experience
Senior ML Engineer, Living Homes
January 2025 – Present
- Senior ML Engineer for the product's cognitive AI backend layer
- End-to-end feature development on multiple critical path features with product owners from user story and design to QA handover.
- Built an assortment of features such as: Constrained conversational smart home agent with tools, RL-based sleep environment optimization, statistics-driven sleep & health analysis, user voice ID recognition, and a dynamic home automation rule engine.
Technologies: Python, Azure, PostgreSQL+MongoDB, Docker, Kubernetes, Temporal, LLMs, Langchain, RL
Senior Data Scientist & Team Lead, VMware Carbon Black
June 2023 – September 2024
- Team Lead for the data team, encompassing data visualization, data engineering, and data science.
- Led collaborations between data teams, engineering, customer success, and product management teams to plan and deliver successful data projects for EDR product cost optimization and customer analytics.
Technologies: Python, AWS, GCP, PostgreSQL, Spark, Tableau
Senior Data Scientist, VMware Carbon Black
Sep 2021 – June 2023
- Senior Data Scientist specializing in the cybersecurity domain at Carbon Black.
- Built signature projects: cloud product cost monitoring and reduction in AWS, EDR ML-driven alerts, cost anomaly detection, LLMs for explainable alerting, and a customer churn model.
- Created cost monitoring tools, anomaly models, and governance methodologies to monitor Carbon Black's AWS & GCP cloud infrastructure costs at a granular level.
Technologies: Python, PyTorch, TensorFlow, AWS, GCP, Terraform, LLMs, Langchain
Senior Data Scientist, Amplify Analytix LTD
Jan 2018 – Aug 2021
- Senior Data Scientist specializing in delivering analytics to sales and marketing teams.
- Product Dev Lead for the first marketable product at Amplify Analytix on optimizing SEO using reinforcement learning.
Technologies: Python, R, Reinforcement Learning, MySQL, Tableau
ML Research Mentor, UCT, Cape Town
July 2015 – Dec 2017
- Machine Learning Research Mentor for the Centre for Artificial Intelligence Research (CAIR) Lab internship program at the CSIR:
- Led research on the feasibility of robotic 3D space understanding, and new time-series forecasting models for financial markets.
Technologies: Python, Machine Learning, Time-Series Forecasting
Projects
Conversational Smart Home Agent
A constrained conversation agent with functions/tools, built for the Living Homes product.
- AI agent that can perform various smart home automations and tasks, conduct sleep and health analysis, guided activities and proactive user environment optimization.
- All done using function calling to internal microservices and external data APIs.
Technologies: Python, Azure, PostgreSQL+MongoDB, Docker, Kubernetes, Temporal, AI, Langchain, Agents
AI-based User Voice ID Recognition
AI-based user voice ID recognition for the Living Homes product.
- Built a voice ID recognition system that can identify users by their voice in near-real-time, and use that to personalize the user's experience.
- Achieved using fine-tuned Nvidia NeMo models.
Technologies: Python, Azure, Deep Learning, AI
EDR ML-Driven & Explainable Alerting
Machine-learning alerting for endpoint detection and response (EDR), with LLM-based explanations to make alerts interpretable.
- Developed ML-driven alerting to improve detection signal in the EDR product.
- Applied LLMs to generate human-readable explanations for alerts.
Technologies: Python, PyTorch, LLMs, Anomaly Detection, Bayesian Methods
SEO Optimization via Reinforcement Learning
Amplify Analytix's first marketable product, optimizing SEO using reinforcement learning.
- Led product development from concept toward a marketable offering.
- Designed the reinforcement-learning approach driving SEO optimization.
Technologies: Python, Reinforcement Learning
Skills
Languages: Python, Java
ML Methods & Stack: TensorFlow, PyTorch, RL, Bayesian Methods, Langchain, LLMs
Cloud & Infra: GCP, AWS, Terraform, Azure, Docker, Kubernetes, Temporal
Data & Databases: MySQL, PostgreSQL, MongoDB, Spark, Redis, Neo4j
BI & Reporting: Tableau, Mode, Looker
Education
Master of Science in Machine Learning and Artificial Intelligence
KU Leuven, Belgium
Bachelor of Science in Computer Science, Physics, and Mathematics
UCT, Cape Town
Languages
English: Native Proficiency
Bulgarian: Native Proficiency
Dutch: Intermediate
Danish: Intermediate
Publications
Gueorguiev, V., & Kuttel, M. (2016, September). Implementation, Validation and Profiling of a Genetic Algorithm for Molecular Conformational Optimization. In Proceedings of the Annual Conference of the South African Institute of Computer Scientists and Information Technologists (p. 16). ACM.