🔸 Toronto, Canada 🔸 Bilingual EN / 中文 🔸 Cambridge DS & ML, 2026 🔸 Available now
Data Science
Data scientist working at the intersection of NLP, time-series forecasting, and behavioural ML. Recently built a hybrid SARIMA-LSTM ensemble that beat standalone deep learning by ~6 points of MAPE on volatile retail demand, and an NLP pipeline that mined 30,000+ customer reviews into 44 actionable dissatisfaction clusters.
I'm completing the University of Cambridge Certificate in Data Science with Machine Learning (2025–2026), with an employer project for the National Institute for Health and Care Excellence (NICE) building a SNOMED CT retrieval-augmented generation pipeline. Other Cambridge work spans hybrid SARIMA-LSTM forecasting on Nielsen BookScan, applied NLP for customer-review mining, anomaly detection on maritime sensor data, and supervised dropout prediction.
Before Cambridge, I spent four years at Kantar Worldpanel Shanghai running quantitative engagements for global beauty and FMCG clients — Shiseido, Estée Lauder Companies, Beiersdorf, Kenvue, Coty. That commercial grounding is why every model I build has a stakeholder-readable narrative attached, not just a metric.
Python, SQL, Excel, Tableau, BioBERT, FAISS, and TensorFlow/Keras round out the stack. Bilingual English / Mandarin, Canadian permanent resident.
MOBILE
GITHUB
Market Research
Four years of quantitative market research expertise at Kantar Worldpanel Shanghai, serving Shiseido, Estée Lauder Companies, Beiersdorf, Kenvue, and others. Presented at National Seminars as Consumer Insight Experts. Now integrating advanced data science to turn raw category data into meaningful foresight — not just descriptive reports.
From 2021 to early 2025 I was at Kantar Worldpanel Shanghai, working across global beauty and personal-care clients including Shiseido, Estée Lauder Companies, Beiersdorf, and Kenvue. I led behavioural diagnostics — price elasticity, buyer outflow, channel migration, cross-tier duplication — that turned ambiguous client briefs into actionable strategy.
I represented Kantar as a seminar speaker at the China Cosmetic Annual Conference (Guangzhou 2024) and Kantar Worldpanel Annual Client Day (Shanghai 2023 & 2024), as the youngest speaker representing vertical market expertise to CMOs of China's top 50 FMCG brands. I also co-authored the 2023 China Beauty Recovery Whitepaper.
I'm now in Toronto, completing a Cambridge DS & ML credential and looking for roles where category depth and methodology rigour both matter.
MOBILE
GITHUB
Data Science
Market Research