Data Analyst at Quartermaster Inc.
Building AI systems, data pipelines, and data products.
I am a data analyst by role and a data scientist by practice.
My work focuses on building end-to-end systems that take messy real-world data and turn it into structured, usable outputs. I am particularly interested in problems that require combining data engineering, machine learning, and product thinking.
Most of what I build follows a simple pattern:
ingest → structure → reason → ship
AI-powered ETL pipeline ingesting and structuring budget data across 3,696 Canadian municipalities.
- Designed for nationwide coverage
- Converts unstructured documents into structured datasets
- Built for @BuildCanada
Privacy-first personal finance platform.
- Universal transaction schema for cross-source normalization
- AI-driven categorization and forecasting
- Deterministic subscription detection
- Designed for local-first and user-owned data
AI agent skill for managing Outlook email, calendar, and contacts via Microsoft Graph API.
Voice AI system for appliance troubleshooting and repair.
- Retrieval over manuals and transcripts
- Step-by-step guided workflows
- Real-time voice interaction
Languages: Python, TypeScript, R, SQL
Data & ML: PyTorch, TensorFlow, scikit-learn, Pandas, NumPy, PySpark
Backend & Infra: Node.js, PostgreSQL, Docker, Linux, GitHub Actions
Analytics: Tableau, Power BI
Tools: Jupyter, LaTeX, Langfuse
- AI agents and tool orchestration
- Retrieval-augmented generation (RAG) systems
- OCR to structured data pipelines
- Observability for LLM systems
- Privacy-focused system design
AI systems, GIS, electric vehicles, and the intersection of software and the physical world
LinkedIn: in/andrewagirgis/
Website: andrew-girgis.com
Email: andrew_girgis@hotmail.com