Portrait of Arian Naseh

Toronto, ON

Arian Naseh

Applied Machine Learning Engineer & Researcher

Senior Machine Learning Engineer with 5+ years of experience designing scalable data pipelines and production-first AI solutions. I build intelligent, integrated systems that simplify complex workflows and drive tangible business value.

Download Résumé Email LinkedIn GitHub Medium Scholar

About

I am a Senior Machine Learning Engineer who treats AI not just as an R&D experiment, but as a core product engine. I have always been driven by the thrill of building things that leave a mark, creating solutions that make people’s lives easier and directly drive top-line growth.

Over the past five years working across fintech startups and high-stakes financial institutions, I have seen firsthand that a highly accurate model is useless if it sits in an isolated silo. I champion a production-first, reformist approach to machine learning. I don't just train algorithms; I architect the end-to-end systems that integrate those models directly into everyday applications. I firmly believe that seamless integration is the only way to ensure user adoption and actually measure business impact.

My work is guided by a pragmatic engineering framework, heavily focused on establishing and standardizing best practices:

  • Simplicity first, iterate second: I focus on establishing a robust, working baseline as quickly as possible. Simplicity doesn't mean cutting corners; it means proving value before introducing complexity.
  • Engineer for integration, not just accuracy: A model is only successful if it fits naturally into existing workflows and architectures.

I thrive in cross-functional environments, whether at a high-throughput tech company or a fast-paced quantitative finance firm, where I can bridge the gap between technical complexity and real-world business value.

Core Competencies

Programming & Data

Python (Primary), Java, SQL, Shell Scripting, Snowflake, BigQuery, Blob Storage

ML & Modeling

Classification, Regression, Ranking, XGBoost, CatBoost, LightGBM, PyTorch, TensorFlow, Time-Series Forecasting

Search & Retrieval

Retrieval-Augmented Generation (RAG), Vector/Hybrid Search (Azure AI Search), Azure OpenAI Embeddings, ChromaDB, DSPy

ML Infrastructure

Online Experimentation, Azure ML, GCP, CI/CD, Drift Detection, Docker, FastAPI, Asynchronous Queuing

Experience

Senior Machine Learning Engineer

Northbridge Financial

May 2022 – Present · Toronto, ON

Search, Retrieval & Intelligent Automation
  • Engineered a RAG-based automated quote generation pipeline for heterogeneous broker submissions, projected to reduce time-to-quote from 1–2 days to under 5 minutes.
  • Designed document-aware chunking algorithms mapping unstructured PDFs to Markdown hierarchies, ensuring semantic cohesion and high-fidelity retrieval.
  • Integrated Azure AI Search and OpenAI embeddings with downstream rating engines, a workflow anticipated to increase quote-to-submission ratios by 3X.
  • Architected structured LLM data extraction via Pydantic validation, establishing a scalable system expected to boost bind-to-quote ratios and drive top-line revenue growth.
  • Automated >70% of First Notice of Loss (FNOL) intake via a LangChain and FastAPI microservice, significantly reducing manual processing bottlenecks.
  • Engineered an asynchronous queuing mechanism to manage Azure OpenAI token limits, ensuring low-latency inference for high-concurrency request workloads.
Ranking, Predictive Modeling & ML Systems
  • Deployed XGBoost frequency and severity models for targeted risk interventions, yielding over $1M in annual operational savings.
  • Designed organization-wide AI monitoring dashboards to evaluate production automation rates, extraction accuracy, and model failure modes.
  • Operationalized automated retraining and drift monitoring pipelines, proactively preventing silent production degradation and ensuring long-term model stability.
  • Built MLForecast and Prophet time-series models with quarterly retraining, securing a $500K resource expansion by directly informing executive workload strategy.
  • Developed robust ETL pipelines utilizing Snowflake and AzureML to integrate and transform large-scale datasets for enterprise ML applications.

Machine Learning Consultant

Docma · Part-Time Contract

Jun 2023 – Jan 2024 · Toronto, ON

  • Optimized high-throughput API endpoints for advertising analytics, significantly reducing latency for real-time campaign dashboards.
  • Refactored audience segmentation models, improving execution efficiency while preserving rigorous statistical fidelity and targeting accuracy.

Data Scientist

TicTie Labs

Jun 2021 – May 2022 · Toronto, ON

  • Engineered supervised machine learning algorithms to predict yield, successfully optimizing environmental controls for indoor farming facilities.
  • Developed real-time SQL and Python data pipelines to track operational metrics, driving data-informed agricultural decisions.

Co-Founder

ETF Ocean

Jan 2021 – Jan 2022

  • Co-founded a fintech startup focused on democratizing algorithmic investment strategies.
  • Built and deployed portfolio selection systems in Python, integrating technical and fundamental financial indicators.
  • Designed ETL pipelines to handle large-scale historical financial data, ensuring accuracy and consistency.
  • Developed modules for validating portfolio back-tests, enhancing strategy robustness and decision-making.

Research Assistant

York University

Sept 2019 – May 2021 · Toronto, ON

  • Applied Deep Reinforcement Learning algorithms to optimize IoT network caching strategies, effectively reducing systemic energy consumption by 30%.

Education

York University

M.Sc. Computer Engineering — GPA: A

Sept 2019 – May 2021 · Toronto, ON

Key Courses: Machine Learning Theory, Probabilistic Models and Machine Learning, Data Mining

Amirkabir University of Technology

M.Sc. Digital Systems — Electronics — GPA: A

Sept 2016 – May 2019 · Tehran, Iran

Key Courses: Statistical Machine Learning, Artificial Intelligence, Computer Vision, Digital Signal Processing

University of Guilan

B.Sc. Electrical Engineering — Electronics — GPA: B+

Sept 2011 – Apr 2016 · Rasht, Iran

Key Courses: Statistics, Algebra, Signal and System Analysis

Publications

  • A. Nasehzadeh and P. Wang, “A Deep Reinforcement Learning-Based Caching Strategy for Internet of Things,” IEEE/CIC ICCC, 2020.
  • H. Wu, A. Nasehzadeh and P. Wang, “A Deep Reinforcement Learning-Based Caching Strategy for IoT Networks With Transient Data,” IEEE Transactions on Vehicular Technology, 2022.

A Little More About Me

  • ☕ Perfecting my coffee brewing techniques, from dialing in the exact espresso shot to experimenting with pour-overs, and hunting for Toronto's best hidden cafes.
  • 🏋️‍♂️ Trying to stay physically active by hitting the gym or getting a few laps in at the swimming pool.
  • 🥾 Co-hosting a bi-weekly summer hike and walk club with my wife. It is a fantastic way to build a sense of community, meet new people, and burn a completely negligible amount of calories!
  • 🐶🐈 Watching an unreasonable amount of dog and cat videos
  • ⚽ Unwinding with a quick game of FIFA.