Hello, I'm Maaheen Yasin

Machine Learning Engineer at MIAL, SFU | Master's in Professional Computer Science

2+ years FinTech Data Engineer · 7+ months Machine Learning Engineer | Canadian Citizen

About Me

Profile Picture of Maaheen Yasin

Hi, I’m Maaheen! I’m a Machine Learning Engineer who’s spent the last few years jumping between FinTech data pipelines and medical imaging AI, and honestly, I’ve loved every bit of it.

I started my career at i2c Inc, a global payments and banking solutions provider, where I went from writing SQL based BI reports to building real-time Kafka data pipelines for large-scale financial data. After two years, I made the leap to grad school at Simon Fraser University, where I dove deep into Machine Learning, NLP, Computer Vision, Data Mining, and Special Topics in AI. That journey brought me to MIAL, the Medical Image Analysis Lab, where I am currently building a PyTorch pipeline for arthritis classification and working on making AI that clinicians can actually trust.

When I’m not doing that, I’m deep in a fantasy novel, painting on canvas, or at the gym. Always happy to connect, feel free to reach out!

Skills

Programming Languages

  • Python
  • C++
Python logo C++ logo

Machine Learning

  • PyTorch, Scikit-Learn
  • Pandas, NumPy, Matplotlib
  • Supervised and Unsupervised Learning
PyTorch Pandas NumPy Matplotlib Scikit-Learn

Data Engineering

  • Apache Kafka, Apache Spark
  • ETL/ELT, SQL
  • Data Warehousing, Data Lakes
  • Data Quality
Apache Kafka Apache Spark SQL Data Warehousing

Natural Language Processing

  • HuggingFace Transformers, LLMs
  • RAG, RAGAS
  • Word Embeddings, Sequence Models
HuggingFace NLP LLM

Computer Vision

  • Image Classification, Image Segmentation
  • Object Detection
  • Domain Generalization, Out of Distribution Detection
  • CNN, GAN, ViT, MobileViT, DINOv2
OpenCV Computer Vision AI

Cloud & DevOps

  • Microservices, Docker
  • Google Cloud Platform
  • Cloud Run, Kubernetes Engine, Compute Engine
Docker Google Cloud Platform Cloud Run Kubernetes Engine Compute Engine

Development Tools

  • Visual Studio Code, Google Colab, Jupyter Notebook
  • Git, GitHub, CI/CD (BitBucket)
  • Jira, PuTTY (Terminal for SSH), Linux
  • Agile/Scrum, SDLC
VS Code Git GitHub Jira Linux

Projects

RepoHero

RepoHero: Retrieval-Augmented Code Understanding System

RepoHero: Retrieval-Augmented Code Understanding System

Team Member (3 Members)

Python, HuggingFace (SentenceTransformer, CrossEncoder), ChromaDB, Ollama, RAGAS

Built a code intelligence system that answers natural-language queries over Python repositories by deeply understanding code structure, not just surface text. Engineered a two-stage retrieval pipeline, AST-aware chunking paired with bi-encoder retrieval and cross-encoder re-ranking, to surface the most semantically relevant code context before generation. Evaluated rigorously using RAGAS across 130 questions on three OSS libraries, pushing accuracy from 73.8% to 80% over the naive baseline.

View GitHub
Recommendation Model Benchmarking

Recommendation Model Benchmarking Across Cold-Start Severity Levels

Recommendation Model Benchmarking Across Cold-Start Severity Levels

Team Member (2 Members)

Python, PySpark (MLlib, ALS), HuggingFace (SentenceTransformer), PyTorch, Pandas, Matplotlib

Built an end-to-end benchmarking pipeline to evaluate how ALS, Semantic Embeddings, and DropoutNet perform as users go from having no interaction history to being fully established, using 1.37M Amazon Books interactions across 594K items. Engineered a PySpark ETL pipeline to ingest, clean, and index the dataset into Parquet, then measured each model's recommendation quality via Recall@100 and NDCG@100 at varying amounts of user history. DropoutNet led across the board, ALS degraded by 68% as users became more established, and Semantic Embedding proved the most stable fallback when no interaction data was available, findings distilled into a model selection guide for production use.

View GitHub
Building Damage Assessment

Building Damage Assessment Using Satellite Imagery

Building Damage Assessment Using Satellite Imagery

Team Member (3 Members)

Python, PyTorch, SAM2, Edge Aware Loss

Trained and optimized the DAHiTrA model to assess building damage from satellite imagery post-disaster. Addressed key challenges like domain generalization, dense urban building merging, and small structure underrepresentation. Integrated Edge Aware Loss (F1: 0.903, IoU: 0.834) and developed a SAM2-based boundary refinement pipeline for improved segmentation in dense areas.

View GitHub
Weather Notification System

Scalable Microservices-Based Weather Notification System

Scalable Microservices-Based Weather Notification System

Team Member (5 Members)

Python (Flask), Docker, Google Cloud Platform, Cloud Run, Kubernetes Engine, Compute Engine, RESTful APIs, BigQuery, Google Cloud Monitoring

Designed and implemented a scalable weather notification system using microservices architecture, with each service containerized in Docker and deployed across Cloud Run, Google Kubernetes Engine, and Compute Engine. Leveraged Google Cloud Monitoring to benchmark performance, analyze resource utilization, and explored trade-offs across serverless, containerized, and VM-based environments.

View GitHub
Image Generation With DDPM

Image Generation With Denoising Diffusion Probabilistic Model

Image Generation With Denoising Diffusion Probabilistic Model (DDPM)

Graduate Coursework

Python, PyTorch, NumPy, Matplotlib, Diffusion Models, U-Net, Deep Learning

Implemented a DDPM using PyTorch based on the original paper Denoising Diffusion Probabilistic Models by Ho et al, designing both forward and reverse diffusion processes and a custom time-conditioned U-Net. Successfully trained on the AFHQ dataset, achieving an FID of 24.43 and demonstrating stable convergence with limited computing resources.

View GitHub
COVID-19 SOPs Compliance System

IoT-Based COVID-19 SOPs Compliance System with Edge AI and Computer Vision

IoT-Based COVID-19 SOPs Compliance System with Edge AI and Computer Vision

Team Lead (3 Members & 2 Supervisors)

Python, CNN, MobileNetv2, YOLO, MediaPipe Library, Embedded Systems, NVIDIA Jetson Nano, Arduino Uno, IoT Sensors

Spearheaded and developed an automated COVID-19 SOPs compliance system using computer vision and IoT sensors for real-time mask detection, social distancing, and vital sign monitoring. Deployed CNN for face mask detection and MobileNetV2 for social distancing monitoring on NVIDIA Jetson Nano, achieving 90.1% accuracy. Recognized among the top 4 of 28 capstone submissions and showcased at SOFTEC 2022, NUCES' premier tech competition.

View GitHub
Personal Portfolio Website

Personal Portfolio Website

Personal Portfolio Website

Individual Project

HTML, CSS, JavaScript

Interested in seeing the behind the scenes of this website? Check out the source code on GitHub!

View GitHub

Professional Experience

Medical Image Analysis Lab (MIAL) at Simon Fraser University

Machine Learning Engineer

Part-time; transitioned from Full-Time Co-op Oct 2025 - Present
  • Architecting an end-to-end machine learning pipeline for arthritis detection across 6+ datasets, adapting a research model to heterogeneous data and optimizing performance to cut diagnostic time by 35%
  • Translated clinical workflows into ML specifications, defining model inputs, outputs, constraints, and interpretable visualizations in collaboration with clinicians and researchers for downstream integration
  • Curated, validated, and standardized 6+ large-scale imaging datasets, defining unified formats, inclusion/exclusion rules, and annotation standards to de-risk model training and evaluation
  • Performed iterative exploratory data analysis on 6+ heterogeneous datasets (30,000+ images), identifying cohort bias, metadata gaps, and annotation inconsistencies to anticipate real-world model failure modes

i2c Inc.

Global FinTech Provider of B2B payment solutions for serving enterprise clients like CIBC and partnering with industry leaders Mastercard and Visa.

Software Engineer (Data Engineering Team)

Jan 2023 – Jul 2024
  • Designed and prototyped a Kafka-based real-time distributed data pipeline (POC) as a scalable replacement for batch ETL, ensuring high data availability for analytics and ML training, reducing end-to-end latency by 83%
  • Collaborated with cross-functional teams (Database, Performance, Cloud, Security, Design Review Board) to define data semantics, functional specs, and production-ready architectures, driving adoption of the new pipeline
  • Mentored and onboarded 2 graduate hires, covering financial reporting, data warehousing, Kafka pipelines, and billing data quality, and authored 4+ recorded training modules on Kafka and data quality to support scalable onboarding

Associate Engineer (Business Intelligence Team)

Jul 2022 – Dec 2022
  • Built and optimized 13+ complex SQL views to ingest, cleanse, and transform large financial datasets for self-serve analytics and decision support, applying FAIR principles and data governance best practices to ensure data accuracy
  • Enhanced data warehouse reliability by building SQL and ETL data quality checks for duplicate and missing records, improving data accuracy and cutting new client onboarding effort by 48% with automated configuration scripts
  • Collaborated with QA to develop and execute 12+ manual unit, system, and regression test plans, supporting test-driven development and continuous integration practices to deliver reliable, secure production releases

Education

Simon Fraser University (SFU)

Master’s in Professional Computer Science - Visual Computing

Sep 2024 - Apr 2026 (Expected)
  • Concentrations: Computer Vision, Deep Learning
  • Coursework: Machine Learning, Visual Computing, Distributed & Cloud Systems, Natural Language Processing (NLP), Special Topics in AI, Data Mining
  • Member of Women in Computing Science (WiCS)

National University of Computer & Emerging Sciences (FAST-NUCES)

Bachelor of Science in Electrical Engineering

Sep 2018 – Jul 2022
  • Concentration: Computer Engineering
  • Coursework: Data Structures & Algorithms, Object-Oriented Programming, Database Fundamentals, Data Communication & Networks
  • Member of NUCES IEEE Society

Awards & Achievements

Silver Medalist for Academic Excellence

  • National University of Computer & Emerging Sciences (FAST-NUCES)
    • Recognized as 1st runner-up in academic performance for Spring 2022 semester.
    • Awarded a silver medal and cash prize.

    Dean’s List Honor Roll Recipient (Spring 2022)

  • National University of Computer & Emerging Sciences (FAST-NUCES)
    • Awarded for achieving a GPA of 3.91 in the Spring 2022 semester.
    • Name published on the official Dean’s List and received a certificate of recognition.

    Champion, Intra-FAST Programming Competition

  • NUCES ACM Society
    • Secured first place in a high-intensity, team-based coding competition using Python.
    • Recognized as “Maestro Coder” for exceptional problem-solving speed and accuracy, and awarded a cash prize.

    Champion, Elementary Robotics Workshop

  • NUCES IEEE Society
    • Engineered a line-following robot using Arduino Uno and infrared sensors.
    • Achieved first place in the workshop’s robotics race, recognized for best speed and precision.

    Extra-Curricular Activities

    Hackathons & Competitions

    Educational Workshops

    Mathematics & Linguistic Contests

    Community Service

    Volunteer Researcher

    Anti-Plastic Ambassador Program (Environment Society)

    Collaborated with a team of 8 to research plastic consumption and its environmental impact, co-authoring a thesis, conducting public surveys, and presenting actionable solutions to promote sustainable policies.

    Community Volunteer

    SOS Saturday's Initiative (Parbat Foundation)

    Led an arts and crafts session at the SOS Children’s Villages, inspiring children through creative activities and decorating the library door with a collaborative snowman art project to foster an inclusive, joyful environment.

    Charity Fundraising Volunteer

    Community Club

    Participated in bake sales and fundraising events, engaging the community to support local charitable initiatives and causes.

    Interests

    Novel Reading

    A big fan of fantasy novels, with favorites including the ACOTAR series, Divergent series, and The Cruel Prince series.

    Canvas Painting of Nature

    Love to paint nature landscapes on canvas with acrylics. A couple of artwork samples are showcased here, with all images protected and not to be reshared or used without permission. Thanks!

    Acrylic Nature Painting 1 Acrylic Nature Painting 2

    Hiking and Nature Walks

    Enjoy spending time outdoors, whether on a hike or a casual walk in nature, to unwind and recharge.

    Let’s Connect

    I'd love to hear from you — whether it's for opportunities, collaboration, or just to say Hi!

    maaheenyasin77@gmail.com

    +1 (647) 918-5674

    github.com/Maaheen11

    linkedin.com/in/maaheen-yasin