Research Topic: Relative Valuation with Machine Learning

Professional Services/AQUIA Scientist Team• Collaborated with 7+ consultants and 6 engineering teams across 4 global locations, delivering 6 quarterly insight reports and independently resolving multiple client-raised issues efficiently.• Directed an end-to-end Machine Learning project using Python and SQL, successfully launching an automated segmentation product that groups millions of users within 1 minute.• Developed and maintained 5+ interactive dashboards with Looker Studio and Python, enabling real-time monitoring and increasing efficiency by 15%.

Corporate Recruitment Strategy: Data-Driven Evaluation Methods and Realistic Outcomes• Utilized Python NLP to process 180 million job descriptions, transforming unstructured text into 7 quantifiable indicators to evaluate recruitment proactiveness, aiding analysis of its relationship with financial performance.• Extracted industry, salary, and competition level from 180 million job descriptions, allowing multi-dimensional analysis of recruitment strategies and their connection to company financial data.

• Cooperated with BENQ product managers to address review process inefficiencies by developing a deep learning-based (BERT) review analysis system, which increased productivity by 20% and provided real-time market insights.• Advised an e-commerce client on customer engagement strategy, developing an AI-powered (GPT model) feedback system that reduced response times by 30% and improved platform engagement.

Shopee project member

• Analyzed over 3 million credit card transactions using Python, creating an effective recommendation system that addressed stagnant repurchase rates, subsequently increasing them by 3.5 times for the sales department. • Conducted a comprehensive brand analysis categorizing 3,500 brands for MOMO, providing clearer market insights that boosted product transaction rate by 30%.