hadi@hadiaghazadeh.com · Calgary, Alberta, Canada
Profile
Senior Data Scientist and applied machine learning researcher with 8+ years building production forecasting, optimization and decision-intelligence systems. I own the end-to-end development and operation of locational marginal price forecasting across thousands of pricing nodes in US and Canadian electricity markets, covering feature engineering, applied research, model development, distributed training, deployment, APIs, monitoring and production maintenance.
Specialized in time-series forecasting, operations research, mixed-integer optimization and reinforcement learning, with hands-on experience building cloud-native ML systems. PhD, published researcher, and author of a practical book on applied reinforcement learning.
Experience
Senior Data Scientist — Enverus
March 2026 – Present · Calgary, Canada
- Own end-to-end development and production operation of locational marginal price forecasting across U.S. and Canadian electricity markets.
- Lead the full ML lifecycle: data analysis, feature engineering, applied research, model development, backtesting, distributed training, deployment, API development, monitoring, incident resolution and ongoing maintenance.
- Build and operate large-scale training and inference pipelines on Kubernetes, with Google Cloud for model serving and storage, Snowflake for analytical workloads and Kafka for streaming data.
- Develop reproducible deployment and model-management workflows with Pulumi and Terraform; production observability with Grafana and Sentry.
Machine Learning Engineer — Bits in Glass
April 2024 – February 2026 · Calgary, Canada (Remote)
- Led technical development of a production hierarchical RAG platform processing more than 20 million oil-and-gas documents, combining structured and unstructured data, hybrid retrieval, embedding-model evaluation, Databricks, LangChain and Model Context Protocol.
- Contributed in a production dynamic fuel-pricing optimization system using integer linear programming and contextual bandits, supporting adaptive margin- and volume-based strategies and cutting pricing decision time from several hours to under 30 minutes.
- Directed implementation of an agentic AI framework for railway legacy-system modernization, automating extraction of business rules and user stories from COBOL applications.
- Built an AI-enabled energy-sector digital twin integrating AWS TwinMaker, IoT sensor data, 3D asset models and natural-language scenario exploration.
- Developed an end-to-end LLM document-processing pipeline with LangGraph and Databricks Mosaic AI, achieving 95%+ structured extraction accuracy on batch invoice data.
- Designed a repeatable evaluation framework for enterprise RAG systems with limited ground-truth data.
Associate Machine Learning Developer — AltaML
May 2023 – September 2023 · Calgary, Canada (Hybrid)
- Contribute in developing a scalable hybrid demand-forecasting system for 100+ fuel-retail locations, improving inventory planning and generating more than $1 million in projected annual savings.
Senior Data Scientist — Snapp
May 2021 – June 2022 · Tehran, Iran (Hybrid) Largest ride-hailing platform in the Middle East
- Led development of a learning-based dynamic surge-pricing system that improved order-fulfilment rate by 5% across more than 200,000 daily orders, improving marketplace balance, revenue performance and driver utilization.
- Directed development of an unsupervised fraud-detection system achieving 90%+ recall and eliminating more than 300 hours of manual review per month.
Data Scientist — Fanap (PARSA)
January 2018 – April 2021 · Tehran, Iran
- Led development of an end-to-end ATM cash-demand forecasting and replenishment optimization system, achieving 17% MAPE and reducing operational costs by 10%.
Education
PhD, Geomatics Engineering — University of Calgary
July 2022 – October 2026 · GPA 4.0/4.0
Research: reinforcement learning, real-time policy optimization, vehicle routing, disaster-response logistics and transportation optimization.
MSc, Industrial Engineering — Amirkabir University of Technology
2015 – 2018 · Tehran Polytechnic
BSc, Industrial Engineering — Iran University of Science and Technology
2011 – 2015
Book and selected publications
Aghazadeh, H. (2026). Applied Reinforcement Learning: Business optimization and LLM fine-tuning. Manning Publications. Publisher
The research page lists all publications with venues and links.
Technical expertise
Forecasting and machine learning — time-series forecasting, demand forecasting, LMP forecasting, feature engineering, backtesting, model evaluation, deep learning, transformer architectures, large language models, RAG, agentic AI, foundation-model fine-tuning
Optimization and decision science — operations research, mixed-integer linear programming, security-constrained unit commitment, power-flow optimization, branch-and-bound, cutting planes, reinforcement learning, deep RL, contextual bandits, dynamic pricing, vehicle routing, supply-chain optimization
Production ML and infrastructure — Kubernetes, Google Cloud Platform, Snowflake, Kafka, Terraform, Pulumi, Grafana, Sentry, Databricks, MLflow, Docker, AWS, SageMaker, Bedrock, Microsoft Azure, CI/CD for ML, model serving, API development, production monitoring
Frameworks and tools — Python, PyTorch, TensorFlow, LangChain, LangGraph, Hugging Face, Model Context Protocol, OpenAI API, Claude API, Apache Spark
Certifications
- Databricks Certified Machine Learning Associate (2024)
- Reinforcement Learning Specialization, University of Alberta / Coursera (2022)
Selected honors and awards
- Alberta Innovates Graduate Student Scholarship (2024) — $31,000 per year for two years
- First Place, Amii Reinforcement Learning Competition, Upper Bound Conference (2024)
- Open Doctoral Scholarship, University of Calgary (2024) — $15,000
- First Place, AI Demand Forecasting Competition, Amirkabir University of Technology (2018) — first among more than 150 teams
Leadership and teaching
- Vice President and Student Representative, Graduate Geomatics Group, University of Calgary (2023–2024) — led a graduate engineering community of more than 50 members
- Teaching Assistant, University of Calgary — Engineering Design, Programming with Data, Spatial Data Mining
- Course instructor and creator — Contextual Multi-Armed Bandits in Python and Reinforcement Learning for Operations Research
- Peer reviewer for machine learning, transportation, optimization and spatial-computing journals
Languages
English (advanced) · Persian (native) · Azeri (mother tongue)