Experience
One internship so far, and it's the reason I moved from full-stack into applied AI.
Data Science Intern, QuantAI
Auckland, New Zealand (remote) · 2025
QuantAI brought me on to help ship three separate applied-AI deliverables for a live client engagement. I worked across the stack: building the retrieval layer for a document Q&A system, tuning a computer-vision pipeline for video analytics, and turning raw sales data into a dashboard executives actually opened.
It's the internship that confirmed applied AI, not just full-stack development, was where I wanted to spend my career. I liked the RAG and CV work far more than I expected to, and I liked shipping something a real client depended on even more.
What I built
Document Q&A system
Built the retrieval layer for an internal document Q&A tool - chunking, embeddings, and a retrieval strategy tuned for accuracy over speed.
YOLOv8 video analytics
Tuned a YOLOv8-based object detection pipeline for a video analytics use case, improving accuracy through frame sampling and confidence thresholding.
Sales analytics dashboard
Cleaned and modelled raw retail-sales data, then built KPI scorecards and regional breakdowns the client's leadership team used directly.
PDF Data Extractor
Built during my internship at Cointe - a Flask app that extracts and validates names, phone numbers, and emails from PDFs, with a Plotly/Dash dashboard and CSV export.
Impact
Owned three deliverables across RAG, computer vision, and BI - end to end, not just a slice of one.
Built for a live client engagement, not an internal sandbox - real deadlines, real feedback.
Went in knowing full-stack development; came out choosing applied AI as the path to specialise in.
Earned AWS and Oracle OCI certifications in Data Science and Generative AI shortly after, to formalise what I'd picked up on the job.
Stack I used
AWS