Hello World!🌍 I am Alireza 🙋‍♂️

About

🔥🚀I am passionate about Data Science (DS), Machine Learning (ML), Artificial Neural Networks (ANNs), Natural Language Processing (NLP), Computer Vision (CV), and Large Language Models (LLMs).

💰📈My mission is to empower my team(s) to tackle complex challenges through advanced analytics and automation. I transform raw data into actionable insights and products, leveraging AI/ML algorithms, statistical models, and data visualizations to reduce costs, promote sustainability, and optimize profitability.

Professional Activities

🤖⚠️At EXP, I lead data- and AI-driven R&D to advance construction automation and energy management. I employ Convolutional Neural Networks (CNNs) to detect construction failures, calculate heat loss from IR thermography images, and estimate window-to-wall ratio for building energy simulations. I also conduct climate risk assessment to explore climate-driven building envelope failures. Collaborating with Canada's National Research Council (NRC), I employ statistical programming for stochastic hygrothermal simulation for moisture prediction and mold risk assessment in buildings. Using NLP and ML, I also automate punch list creation for non-compliance and deficiency tracking in building commissioning and new construction, significantly reducing manual effort.

📊💰With my other code developments, I assist upper management in automating client invoicing, project management, identifying budget surpluses/deficits, tracking employee utilization, and forecasting profitability.

🧠💡At Scale AI, I evaluate LLM performance for coding-intensive roles, as I contribute to projects like Beagle Coding, Coders Full Stack, and Observation Concrete. I contribute to reinforcement learning with human feedback (RLHF), fine-tuning LLM responses, ensuring adherence to accuracy, conciseness, and security standards, and verifying APIs to prevent hallucinations.

👤🏷️At Telus Digitals (formerly Telus International AI), I provided high-quality labeled data for tasks like Named Entity Recognition (NER) and Region of Interest (ROI) annotation, supporting CV and LLM training. This work also established Human-Level Performance (HLP) benchmarks for robust AI evaluation.

🏠💨At UofT, I improved indoor air quality (IAQ) and sustainability by ML and data analytics. Using AI/ML, I predicted HVAC operations based on temperature and humidity changes, forcasted thermal comfort in multi-unit residential buildings (MURBs), and introduced Rapid Quantitative Filter Forensics (RQFF) to expedite airborne contaminant analysis, enabling efficient post-field HVAC filter forensics and laboratory coordination🔍.

💵🩺 My other activities extends to finance, retail, healthcare, and beyond, where I’ve worked on projects like fraud detection, sales optimization, customer churn prediction, breast cancer tumor detection, sentiment analysis, machine translation, self-driving cars, and sports analytics. I also specialize in MLOps, Big Data, and recommender systems, delivering tailored solutions across sectors.

Skills and Experience

Programming (Python, SQL, VBA, C/C++)

DS, ML, & Deep Learning (Pandas, Numpy, Scikit-Learn, TensorFlow, OpenCV, CuDF, XGboost, Polars)

Plotting & Visualization (Matplotlib, Seaborn, Plotly, Pandas, Bar Chart Race)

Text Mining & NLP (nltk, SpaCy, TextBlob)

Statistics (Scipy, StatsModels)

MLOps & Cloud (Docker, FastAPI, Flask, GCP, AWS, DataBricks)

Big Data (PySpark, Spark SQL, Polars, CuDF)

RDBMS (MS SQL Server)

Climate Change & Environment (MeteoStat, PyThermalComfort)