Amir Feizi Email & Phone Number

Canada

Toronto, Ontario, Canada

• Developed several prediction models and ML algorithms for contact centre KPIs across various regions and client segments• Built a modular forecasting framework to optimize contact centre workforce capacity to handle 5 million calls yearly• Created an XAI mix model to interpret and visualize predictions, enhancing transparency in decision-making processes• Conducted large-scale analysis and evaluated several key metrics to provide deeper insights for stakeholders• Validated models using various techniques, graphs, calibration & comprehensive testing scenarios to ensure robustness• Optimized SQL queries to ensure efficient big data processing and reduced computational costs• Developed sentiment analysis on customer call transcripts to extract actionable insights during the pandemic• Proactively initiated analytical solutions, led projects and junior team members, collaborated with cross-functional teams and articulated ideas clearly

Montreal, Quebec, Canada

• Conducted research on power failure detection (predictive maintenance) of IoT edge devices utilized in smart bus stations • Created a series of intuitive dashboards and reports using PowerBI to extract insight from telemetry sensor data, weather data and real-time public transportation data.• Developed deep learning-based models to forecast power failure and improve IoT edge device battery longevity by 15%• Deploy and automate ML models using MLOps, creating end-to-end pipelines to meet performance goals. Leverage Databricks and Azure services like Data Factory and ML Studio, plus Docker, TensorFlow, and Airflow, to scale and manage production-ready ML solutions effectively.

Montreal, Quebec, Canada

• Conducted research and developed a mathematical model to optimize resource allocation in a multi-agent wireless system.• Proposed a decomposition-based approach to solve nonlinear resource allocation challenges.• Designed a solution for a mixed-integer nonlinear optimization problem, integrating metaheuristic algorithms (Tabu Search, genetic, ..) with Sequential Quadratic Programming (SQP).• Developed an efficient resource allocation algorithm leveraging reinforcement learning principles.

Doctor Of Philosophy - Phd, Computer Engineering

Secrecy-Optimized Resource Allocation for Device-to-Device Communication Undelaying Cellular Networks

Master'S Degree, Electrical, Electronics And Communications Engineering

Power Control and Cooperative Relay Selection in Cognitive Radio Networks