Click & Drag to create dust
TAP TO TOGGLE DUST

Noah Oosting 

Junior Developer & AI Engineer

Professional Overview

👋 Professional Profile

Aspiring software developer bridging the gap between traditional engineering and modern AI orchestration. With a strong foundation in Android, Web, and Game development, I am currently specializing in building agentic workflows and intelligent systems. Passionate about writing clean, maintainable code and solving complex problems through innovative automation.

Core Technologies

PythonJavaTypeScriptNext.jsReactTensorFlowKerasPyTorchPyTorch LightningspaCyTransformers (BERT, GPT-3)PandasScikit-learnSpring BootJakarta EEOracle PL/SQL

Experience

AI Development Student

2025 - Present
Fanshawe College
  • Designing and deploying advanced deep learning architectures including CNNs, RNNs, and GANs using TensorFlow and PyTorch.
  • Implementing state-of-the-art Natural Language Processing solutions utilizing Transformer architectures (BERT, GPT-3) and spaCy for intent recognition.
  • Analyzing and mitigating machine learning security vulnerabilities, developing robust models resilient against data poisoning and evasion attacks.
  • Executing end-to-end data mining and visualization pipelines to extract actionable insights from large-scale datasets.
  • Optimizing model performance by leveraging NVIDIA GPU hardware acceleration systems.

Education

Artificial Intelligence and Machine Learning

Fanshawe College
Post-Graduate Certificate (AIM)
2025 - 2026

Machine Learning Optimization Strategies

INFO-6154
Model Compression

Pruning and quantizing VGG16 models to reduce size by 50% while maintaining 95% accuracy for edge deployment.

GPU Benchmarking

Profiling deep CNN training loops to identify bottlenecks and demonstrating 10x speedups via CUDA acceleration.

NVIDIA CUDATensorRTOptunaPyTorch Profiler

Deep Learning with Tensorflow & Keras 2

INFO-6152
Generative Adversarial Networks

Building DCGANs to synthesize realistic facial imagery and implementing Neural Style Transfer.

Sentiment Classification

Designing LSTM/GRU recurrent networks for high-accuracy sentiment analysis on the IMDb dataset.

TensorFlow 2.xKeras Functional APIGANsOpenCV

Natural Language Processing 2

INFO-6153
BERT Fine-Tuning

Adapting pre-trained BERT models for specialized financial sentiment analysis tasks.

Creative Text Generation

Implementing autoregressive generation pipelines using GPT-2 for creative fiction.

Hugging FaceBERTGPT-2PyTorch

Machine Learning Security

INFO-6149
Adversarial Defense

Generating FGSM adversarial examples to fool classifiers and implementing defensive distillation to restore robustness.

Data Poisoning Detection

Simulating training set contamination and building outlier detection systems to sanitize data.

Adversarial Robustness ToolboxCleverHansPython

Social Media Analytics

INFO-6155
Brand Sentiment Monitor

Real-time ingestion and sentiment classification of live social streams for brand reputation tracking.

Twitter APINetworkXNLTK

Deep Learning with Pytorch

INFO-6147
Custom Image Classifier

Training CNNs on custom collected datasets using Transfer Learning (ResNet) and data augmentation.

PyTorch LightningTorchvisionTransfer Learning

Computer Programming

Algonquin College
Ontario College Diploma
2020 - 2022

Enterprise Application Programming

CST8277
Enterprise REST API Service

Engineered a scalable backend service using Spring Boot, designing RESTful endpoints and implementing role-based security protocols.

Java EESpring BootJPA/HibernateREST APIs

OOP with Design Patterns

CST8288
Legacy Code Refactoring

Refactored monolithic Java applications by applying Singleton, Factory, and Strategy patterns to improve maintainability.

JavaDesign PatternsSOLIDJUnit

Advanced Database Topics

CST8276
Hybrid Database Solution

Developed complex PL/SQL stored procedures for business logic automation and integrated NoSQL data streams.

Oracle PL/SQLStored ProceduresMongoDBData Warehousing

Mobile Graphical Interface Programming

CST2335
Native Android Productivity App

Implemented complex UI layouts, local data persistence, and asynchronous network operations.

Android SDKKotlinSQLiteFragments

Featured Projects

GovCan Scraper Pipeline
GovCan Scraper Pipeline

A sophisticated data pipeline for scraping and processing Canadian government documents for RAG systems. Scrapes CRA and Service Canada documents, extracts text from PDFs, and generates vector embeddings.

PythonSeleniumPDFPlumber
AdultingOS V2
AdultingOS V2

A 'Government Document Assistant' designed to simplify complex Canadian forms. Features a wizard-style interface, plain language translations, and AI-driven form assistance to help users navigate bureaucracy.

FastAPIDjangoReact (Vite)
goblin-archivist-dev
goblin-archivist-dev

A Godot-based prototype game demonstrating real-time LLM integration and agentic gameplay via a custom TCP middleware bridge. Relays commands between an external MCP client and the game environment.

Godot 4GDScriptPython
Agile Capstone Client Project
Agile Capstone Client Project

Lead Developer for an external client project delivering a custom software solution within a strict Agile environment. Managed the full SDLC from requirements gathering to deployment.

JavaSpring BootAgile/Scrum
RuneLite Plugin Suite
RuneLite Plugin Suite

A collection of performance and audio enhancement plugins for the Old School RuneScape client. Includes the 'Pokemon Low HP' plugin which plays anxiety-inducing audio cues when health drops below a threshold.

JavaRuneLite APIMaven
📂

View Project Archive

6 more projects...

❤️

Visit London Ontario Charity Site

Curated and Created by Me