Yik Sum Man

Have practical experience in the application of state-of-the-art AI techniques, including embeddings and Large Language Models, in real-world business scenarios at GPA Global. As an AI and Computer Science student at the University of Edinburgh, holds a comprehensive understanding of various machine learning models, such as supervised, unsupervised learning, and transformers. Demonstrates a commitment to staying abreast of the latest trends in data science and AI through regular engagement with academic papers, podcasts, and industry newsletters.

Project 1: SectionSeeker AI - Extended Proof of Concept at GPA Global

• Designed SectionSeeker AI to rapidly retrieve information across multiple documents, efficiently extracting and presenting the most relevant sections directly to the user
• Integrated OpenAI's embedding models to analyze documents and interpret user queries, applied GPT while processing documents, innovatively adapting LLMs for business tasks.
• Addressed generative AI's hallucinations issue, significantly improving decision-making efficacy.

Project 2: Deep Dive into Machine Learning Models

• Developed a comprehensive understanding of machine learning by constructing and analyzing various models from scratch using Python.
• Explored Supervised, Unsupervised, Ensemble models, and Transformer architectures.
• Focused on state-of-the-art algorithms and practical applications for continuous skill enhancement.
  

The Data Science Course: Complete Data Science Bootcamp

Udemy

Learn SQL Course

Codecademy

Key Skills:

• Analytical: Robust logical and data-driven analytical thinking.
• AI Design: Skilled in creating and implementing AI tools.
• Continuous Learner: Committed to staying updated with AI advancements.
• Team Collaboration: Proven teamwork at GPA Global on AI projects.
• Problem Solving: Strong aptitude in tackling complex code challenges.
• Work Ethic: Prioritizes efficiency and impact in a dynamic environment.

Technical Proficiencies:

• Languages: Proficient in Python; familiar with Java.
• Frameworks: Hands-on experience with TensorFlow and Keras; introductory familiarity with PyTorch.
• Libraries: Experience with Scikit-learn, seaborn, Numpy, Pandas, Hugging Face, and OpenAI.
• APIs & Bots: Experience with chat bot-type APIs, including ChatGPT and CodyAI.
• Prompt Engineering: Experienced in refining CodyAI prompts for optimal user interactions.
• ML Focus: Hands-on with application-centric Large Language Models.

ML & DL Algorithms:

• Supervised: Decision Tree, KNN, Linear Regression, Logistic Regression, Naive Bayes, Random Forest, AdaBoost.
• Unsupervised: PCA, K-Means, SVM.
• NLP: Word Embeddings, Tokenization.
• Deep Learning: Neural Networks, Encoder-Decoder, Attention Model, Transformer, LSTM.