Hey, I'm Amirali.
I build reliable data and machine learning systems for messy, real-world technical data.
My work focuses on Python, SQL, and pipeline design: ingestion, cleaning, validation, transformation, and analytics-ready delivery. With an electrical engineering background, I connect system-level domain knowledge with practical data, ML, and AI solutions.
Explore the selected work below, or reach out if you want to connect.
Graduate Studies, Electrical Engineering
Montréal, Quebec, Canada
Focused on research workflows involving IoT and time-series datasets, Spark and SQL transformations, data validation, machine learning support, and reproducible experimentation.
B.Sc., Electrical Engineering
Tehran, Iran
Built the technical base for signal processing, control systems, circuit design, numerical methods, programming, and applied analytical problem solving.
A compact view of my professional experience first, followed by personal project work arranged chronologically.
2024 - Present
2021 - 2024
2018 - 2020
Built an end-to-end pipeline for large-scale embedding datasets with orchestration, feature engineering, dimensionality reduction, and artifact generation.
GitHubIntegrated RAG pipelines, APIs, and workflow automation to classify Toronto building permits and generate structured project outputs.
Developed a scalable image-generation pipeline using Stable Diffusion 1.5 and XL, optimizing CUDA workflows and parallel processing.
Designed a real-time NLP pipeline for social media sentiment during disaster events, enabling live monitoring through interactive dashboards.
Developed Monte Carlo value estimation models for episodic environments, analyzing convergence behavior and variance against dynamic programming baselines.
GitHubImplemented SARSA and Q-learning for stochastic control problems with dynamic reward functions and policy performance benchmarking.
GitHubBuilt and trained a multi-layer perceptron on MNIST, improving performance through hyperparameter tuning, backpropagation, and evaluation workflows.
Designed a decentralized estimation system using Bayesian modeling and distributed Kalman filtering for real-time prediction across multiple agents.
Processed and denoised audio and image signals using filtering techniques and windowing methods to improve signal fidelity and reconstruction quality.
Implemented PID, Lead, and Lag controllers on microcontroller platforms, validating response through oscilloscopes, multimeters, and function generators.
Developed a two-layer PCB for stable 12V and 5V voltage regulation, validating performance through SPICE simulation, thermal checks, and load testing.
Recent coursework focused on analytics, machine learning foundations, and production-style ML workflow design.
Covered exploratory data analysis, dashboards, SQL queries, and predictive modeling using Google tools and open-source libraries.
Completed practical training on forward and backward propagation, activation functions, gradient descent, and tuning multi-layer neural networks.
Designed ML pipelines using Azure visual tools, including preprocessing, model selection, deployment, and performance tracking.
Tools and methods I use to build reliable data products, analytics workflows, and machine learning systems.
