Hey, I'm Hetang.

About.

Projects.

1. 01

   Energy Copilot

   In progress: an AI copilot that answers plain-English questions over public Alberta well production data.

   Learn more >: Energy Copilotsource code: Energy Copilot

   Python - FastAPI - PostgreSQL - pgvector - Next.js

2. 02

   EMG Viability Classifier

   An ML pipeline that classifies prosthetic trial candidate viability from EMG sensor data, built for Alberta BIONIX.

   Learn more >: EMG Viability Classifiersource code: EMG Viability Classifier

   Python - Scikit-learn - NumPy - Pandas - Signal Processing

3. 03

   Batch Normalization Reproduction

   Reproduction of Ioffe & Szegedy's ICML 2015 Batch Normalization paper, validating its core claims on MNIST and CIFAR-10.

   Learn more >: Batch Normalization Reproductionsource code: Batch Normalization Reproduction

   Python - PyTorch - NumPy - Matplotlib

4. 04

   Event Lottery System

   An Android application that enables fair, lottery-based registration for high-demand community events.

   Learn more >: Event Lottery Systemsource code: Event Lottery System

   Java - Android Studio - Firebase

5. 05

   Multithreaded Concurrent Server

   A thread-pool TCP server in C handling concurrent read/write requests with data consistency guarantees.

   Learn more >: Multithreaded Concurrent Serversource code: Multithreaded Concurrent Server

   C - POSIX Threads - Sockets - Concurrency

6. 06

   ATS Resume Analyzer

   A GenAI-powered full-stack web application that evaluates resume compatibility with job descriptions.

   Learn more >: ATS Resume Analyzersource code: ATS Resume Analyzerlive project: ATS Resume Analyzer

   Next.js - TypeScript - React - Google Gemini AI

7. 07

   Boxed

   A cloud storage infrastructure enabling secure file uploads, retrieval, and access control.

   Learn more >: Boxedsource code: Boxed

   React - TypeScript - Node.js - Express - Auth0 - Cloudflare R2 - Docker

8. 08

   Chess.app

   Developed an iOS application using Swift and Xcode that allows a user to play a 2-player game of chess with a timer.

   Learn more >: Chess.appsource code: Chess.app

   Swift

Experience.

1. JAN - AUG 2024; JAN - AUG 2025; MAY - DEC 2026

   Calgary

   Canadian Natural Resources Limited, CNRL

   Systems Developer Student

   • Improved productivity by 30% for 500+ exploitation engineers by designing and shipping a PWA that gives them one reliable source for well production data (C#, .NET, Blazor, REST APIs, Azure CI/CD).
   • Led development of the Enersight integration, a multi-stage import wizard in the capital budgeting app: stages external asset data in Oracle, lets users map, merge or exclude assets against budget entities, and loads only after privilege, version-lock and business-rule validation, with a per-asset audit trail.
   • Cut worst-case load time by 81% (117s → 22s) on a performance-critical VB.NET/ASP.NET and Oracle data path by profiling it, removing redundant fetches and moving a 59s operation off the critical path; added reusable timing instrumentation.
   • Took over as lead developer of the budgeting application after a team transition: sole technical contact for 5 business stakeholders, owner of deployments, and onboarding lead for a new developer; set up twice-weekly standups and a weekly business review.
   • Cut economics regression testing from 10 days of manual UAT to 4 minutes by building a library of 300+ automated unit tests and validation pipelines in C# and VB.NET, saving $60K.
   • Reduced defects by 33% and recovered 100+ hours/yr of engineer idle time by redesigning Oracle SQL schemas, optimizing queries, and enforcing data integrity constraints.

   VB.NET - Oracle PL/SQL - DevOps - Microsoft Excel - Blazor - Azure - ASP.NET - Enersight API - C# - .NET - REST APIs

2. SEP 2025 - Present

   Edmonton

   Alberta BIONIX - School Club

   Machine Learning Team Lead

   • Lead the ML team building control models for a prosthetic knee, managing and mentoring 2 junior engineers; researching reinforcement learning to auto-tune the knee motor's PID controller.
   • Processed and analyzed a 26GB EMG gait dataset using Python (Pandas, NumPy, Scikit-learn), building feature engineering pipelines to extract biomechanical signals used for prosthetic knee control.
   • Developed and evaluated machine learning models (Random Forest, SVM) to classify prosthetic trial participant viability from EMG activation patterns.
   • Designed and trained CNN models for real-time knee flexion prediction, achieving 87% accuracy across locomotion modes (sitting, standing, walking, running).

   Python - Pandas - NumPy - Scikit-learn - CNNs - Random Forest - SVM - Reinforcement Learning

3. SEP 2023 - DEC 2023

   Fort McMurray

   Suncor Energy

   Data Analyst Intern

   • Developed 3 Python-based automation tools for SAP workflows (data maintenance, updates, manipulation), saving 30+ hours/week of manual work and $26K annually for planners and schedulers.
   • Cleaned and processed large datasets using Python, Excel, and VB to ensure data quality, then built analytical dashboards in Power BI to support data-driven decisions across the team.

   Python - SAP - PowerBI - Microsoft Excel - VB

4. May 2019 - SEP 2023

   Calgary

   YMCA

   Aquatics Lifeguard and Swimming Instructor

   • Trained four lifeguards and swimming instructors on effective scanning and teaching techniques
   • Utilized lifesaving techniques to ensure the safety of patrons during a swim test, lessons, on deck and when assisting co-workers
   • Ensured patrons aged from 12-months to 75 years old received an amazing learning experience in lessons
   • Performed regular headcounts, three water tests per day, change room checks, and safety checks of equipment