Herbert Ho Bun Lau
Lead Data Scientist @Capgemini
Signup · Get unlimited contacts
WORK HISTORY
Lead Data Scientist @Capgemini
Design, develop, and deploy AI and machine learning solutions that turn structured and unstructured data into actionable business insights. Work closely with cross-functional stakeholders to deliver production-ready AI applications used in real-world environments.Hands-on development of Generative AI applications and machine learning models to drive measurable business value.Design and deliver multiple RAG (Retrieval-Augmented Generation) chatbot solutions, including prompt engineering, file search capabilities and metadata filtering for improving relevance and performance.Lead development of context ingestion pipelines and metadata generation, significantly improving RAG accuracy and response quality.Collaborate with product, engineering, and business teams to align AI solutions with strategic objectives.Contribute end-to-end in an agile development environment, from experimentation and prototyping to deployment of production-grade AI systems.
EDUCATION
University of Cambridge
Doctor of Philosophy (PhD), Astrophysics
University of Cambridge
Master of Science (MSc), BA, Natural Sciences- Astrophysics
SKILLS
ABOUT HERBERT HO BUN LAU
Hands-on AI consultant/data scientist with more than ten-year experience on machine learning, data analysis, statistics, and applied mathematics. Experience in generative AI, Natural Language Processing (NLP) and classical machine learning such as anomaly detection and forecasting.I train Large Languages Model (LLM) with private data for specific use case such as customised chat-bot.I work on Natural Language processing tools such as sentiment analysis, words embeddings for classification problems based on textural and qualitative data. I have also worked on taxonomy and knowledge graphs.I have experience in various machine learning and statistical data analytic techniques such as Neural network, Bayesian network, Markov-Chain Monte Carlo, A/B testing, Naïve Bayes classification, Natural Language Processing. I work independently on small project as well as working in a team focusing on bigger picture and problems. I have experience cleaning bad data and extracting useful information from large dataset with uncertainty. I am very comfortable with statistics. I work on hypothesis testing to compare theoretical models with observations and use Bayesian statistics to find the most likely scenario.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.