Summary
Overview
Work History
Education
Skills
Languages
Certification
Timeline
Additional Information
background-images
Amir Barzegar

Amir Barzegar

Toronto,Canada

Summary

Experienced Machine Learning Engineer and Senior Data Scientist with over 8 years of expertise in building and deploying real-world, end-to-end machine learning systems across various industries. Specialized in creating predictive models for workforce optimization, logistics, healthcare, and cybersecurity. Proven background in Generative AI, Large Language Models (LLMs), NLP applications, and model fine-tuning. Proficient in designing data pipelines, feature engineering, model development, and deployment of machine learning systems into production environments using scalable cloud-based architectures. Strong expertise in production monitoring, model drift detection, and ML lifecycle management with tools like DVC and MLflow. Experienced in fine-tuning LLMs for region-specific adaptation to provide context-aware and domain-specific answers.

Overview

12
12
years of professional experience
1
1
Certification

Work History

Senior Data Scientist

Airudi Company
06.2021 - Current
  • Developed a workforce prediction and allocation system for port logistics, optimizing task assignments and employee distribution through operations research and machine learning models.
  • Led a forecasting system to predict call volumes, priorities, and workload for 24-hour and long-term planning, using both internal and external factors for better service delivery.
  • Designed and implemented Generative AI models to predict employee safety risks in industrial environments, providing preventive measures and notifying management of potential hazards.
  • Built NLP-driven solutions for information extraction and safety risk prediction, applying LLM fine-tuning, embeddings, and generative model techniques aligned with retrieval-augmented decision processes.
  • Full responsibility for the end-to-end process: data collection, preprocessing, feature engineering, model creation, deployment on cloud (Azure), and exposing predictions via RESTful APIs for integration with backend and frontend systems.
  • Designed and maintained a DVC-based pipeline for model development, experiment tracking, and reproducible training workflows.
  • Integrated models with self-hosted MLflow servers for experiment logging, performance monitoring, and model registry management.
  • Continuously monitored model performance post-deployment, implemented drift detection mechanisms, and conducted model retraining based on project requirements to ensure system reliability and performance improvements.

Research Assistant (RA) – Privacy Preservation in Machine Learning

Polytechnique Montreal, University of Montreal
09.2018 - 06.2021
  • Conducted research on privacy-preserving machine learning techniques, focusing on secure and decentralized training methods for machine learning models.
  • Worked on developing privacy-preserving algorithms that could anonymize sensitive data while maintaining model accuracy and effectiveness.
  • Explored federated learning and differential privacy approaches to ensure that personal data could be used without compromising privacy.
  • Collaborated with faculty and researchers to publish findings in academic journals and conferences, advancing privacy techniques in ML models.

Machine Learning Engineer & Software Developer

Dana Company
01.2014 - 04.2018
  • Developed a machine learning-based anomaly detection system to detect malicious network activities and block malicious nodes from connecting to the server, significantly improving network security.
  • Implemented and optimized the pattern recognition system to continuously monitor network activities and detect any signs of cybersecurity threats.

Education

Master of Science - Machine Learning and Software Engineering

Polytechnique Montreal, University of Montreal
01.2021

Master of Science - Artificial Intelligence

SRB University
01.2013

Skills

  • Proficient in Python, C, and Java
  • Machine Learning: Supervised and unsupervised learning, Generative AI, NLP, embeddings, LLMs, model fine-tuning, PEFT, LoRA, RAG (retrieval-augmented generation) concepts, region-specific LLM fine-tuning for tailored responses
  • Cloud & DevOps: Azure (Azure ML, CI/CD, Data Lakes, Queues), TGI servers, Cloud deployments
  • Data Engineering: Data pipelines, feature engineering, data preprocessing, ETL processes, distributed data processing
  • Data Science: Forecasting, anomaly detection, predictive modeling, operational research, model drift detection, model monitoring
  • MLOps & Lifecycle: DVC pipelines, MLflow for experiment tracking, monitoring, and production deployment decisions
  • Database Technologies: SQL, Redis, Postgres, Cosmos
  • Software Development: Agile, Kanban, Sprint Planning
  • Other: Model deployment as REST endpoints, workforce orchestration, safety risk prediction
  • Tools: NLTK, SpaCy, TensorFlow, PyTorch, Scikit-learn, Docker

Languages

English (Fluent)
French (Intermediate)
Farsi (Native)

Certification

Deep learning + Reinforcement learning

Google cloud attention mechanism

Google cloud Encoder-Decoder Architecture

Google cloud LLM

Google cloud Generative AI Studio

Timeline

Senior Data Scientist

Airudi Company
06.2021 - Current

Research Assistant (RA) – Privacy Preservation in Machine Learning

Polytechnique Montreal, University of Montreal
09.2018 - 06.2021

Machine Learning Engineer & Software Developer

Dana Company
01.2014 - 04.2018

Master of Science - Artificial Intelligence

SRB University

Master of Science - Machine Learning and Software Engineering

Polytechnique Montreal, University of Montreal

Additional Information

  • Familiar with large-scale machine learning system architectures and cloud-based AI models
  • Extensive experience with ML lifecycle management using DVC and MLflow for model tracking, deployment, drift detection, and retraining strategies
  • Passionate about continuous learning and innovation in the field of AI and machine learning
  • Experience working with cross-functional teams in Agile and Kanban environments
  • Strong alignment with Generative AI, NLP, and RAG concepts for building advanced AI solutions in production settings
Amir Barzegar