CS grad building intelligent systems and data pipelines that ship.
I build production LLM systems with LangGraph, AWS Bedrock, and OpenAI — multi-agent pipelines, RAG architectures, and FastAPI backends deployed to AWS ECS Fargate. On the analytics side: Python pipelines, PostgreSQL, SQL, and Power BI dashboards that turn raw data into institutional-grade insights.
Selected projects
InsightStream
A production-grade multi-agent pipeline using LangGraph with 6 specialized agents orchestrating Claude Sonnet 4.5, Llama 3.3 70B (AWS Bedrock), and GPT-5.1/5.2 (OpenAI). Monitors 41 RSS feeds, processes 400+ articles per run with cosine-similarity semantic deduplication and diversity-aware ranking across 8+ categories. Delivers personalized AI-curated daily digests based on natural language preferences. Deployed on AWS ECS Fargate with GitHub Actions CI/CD — 5-minute end-to-end runtime.
Neural Semantic Job Search
LLM-powered resume-to-job matching using Llama 3.3 via AWS Bedrock. Hybrid ranking pipeline beats keyword search.
FinTech RAG Copilot
RAG system answering financial regulatory questions (OSFI docs) with citation-backed responses via ChromaDB + Amazon Titan.
IEEE-CIS Fraud Detection
XGBoost classifier hitting 94.08% ROC-AUC on the IEEE-CIS dataset. Outperformed LightGBM, Random Forest, and Logistic Regression.
Toronto Airbnb Price Prediction
XGBoost regression model achieving R²=0.79 on Toronto Airbnb listings. Feature engineering across location, host tenure, and availability signals.
Financial Market Intelligence & Risk Analytics Platform
End-to-end financial analytics platform ingesting 5 years of daily data for 12 tickers across Technology, Financials, and ETFs. Python pipeline with yfinance & FRED API, 8 institutional-grade risk metrics, 4 PostgreSQL tables with 7 hand-written SQL views, and a 3-page executive Power BI dashboard. Key insight: NVDA returned 1,257% over 5 years with the highest 90-day rolling volatility at 34.8%.
The full stack
From multi-agent LangGraph pipelines and vector databases to Power BI dashboards and AWS ECS Fargate deployments — across AI engineering and data analytics.
AI / ML
- AWS Bedrock (Claude, Llama, Titan)
- OpenAI API (GPT-5.1/5.2)
- LangGraph & Agentic AI
- LangChain / LlamaIndex
- RAG Pipelines
- Semantic Search & Deduplication
- XGBoost / LightGBM / scikit-learn
- Prompt Engineering
- ChromaDB / FAISS
- Hallucination Mitigation
Backend & Cloud
- FastAPI
- REST APIs
- Pydantic Validation
- Async API Handling
- AWS ECS Fargate / ECR
- Docker & Docker Compose
- GitHub Actions CI/CD
- Git & GitHub
- Streamlit
- Vercel / Render
Languages & Tools
- Python (primary)
- Java
- SQL
- Pandas & NumPy
- Matplotlib / Seaborn
- Jupyter Notebooks
- PyPDF
- Linux / CLI
Data & Analytics
- Power BI
- Microsoft Excel (Advanced)
- SQL (Data Analysis)
- PostgreSQL
- Pandas (Data Wrangling)
- Exploratory Data Analysis
- Feature Engineering
- Data Visualization
- Statistical Modelling
- Risk Analytics
- Anomaly Detection
- Regression & Classification
- Dashboard Design
- yfinance / FRED API
Education & trajectory
BSc Computer Science
York University
Honours degree in Computer Science covering machine learning, data structures, algorithms, databases, and software engineering. Graduating 2026.
Agentic AI for Developers
LinkedIn Learning
Concepts and enterprise application of agentic AI systems — designing autonomous agents, tool use, memory, and multi-agent orchestration patterns.
Build REST APIs with FastAPI
LinkedIn Learning
Building production REST APIs with FastAPI — routing, Pydantic validation, async handlers, dependency injection, and API documentation.
Learning Git and GitHub
LinkedIn Learning
Version control fundamentals — branching strategies, pull requests, CI/CD integration, and collaborative workflows on GitHub.
Generative AI with Large Language Models
DeepLearning.AI
Andrew Ng's flagship GenAI course covering transformer architectures, fine-tuning, RLHF, and production deployment practices for large language models.
ChatGPT Prompt Engineering for Developers
DeepLearning.AI × OpenAI
Advanced prompt engineering techniques — systematic prompt design, chain-of-thought reasoning, and LLM application patterns with Andrew Ng.
AI & ML Bootcamp
Udemy
Comprehensive machine learning bootcamp covering supervised/unsupervised learning, neural networks, and end-to-end ML project workflows.
Self-Taught AI & Analytics Trajectory
Independent
ML foundations → data analytics (Excel, Power BI, SQL, PostgreSQL) → GenAI APIs → RAG architectures → Agentic AI with LangGraph → Production on AWS ECS Fargate. 6 projects shipped.
Active member of the Toronto Machine Learning Society and participant in events hosted by the Vector Institute. Staying close to the frontier of applied AI research and connecting with the local ML engineering community.
Open to the right opportunity
I'm a Junior AI Engineer and Data Analyst who builds production-ready systems across both lanes. On the AI side: LangGraph, AWS Bedrock, and OpenAI — multi-agent pipelines, RAG architectures, FastAPI backends on AWS ECS Fargate. On the analytics side: Python, PostgreSQL, SQL, and Power BI dashboards that turn raw data into institutional-grade insights.
Based in Toronto. Open to hybrid or remote Junior AI/ML Engineer and Data Analyst roles in Canada or internationally. Whether you're building with LLMs or need someone to turn data into decisions — let's talk.