AI Data Scientist

Millennium; Hong Kong; On-site; HKD 600,000 - 900,000; Full time

Job summary

Millennium is seeking an AI Data Scientist to work directly with portfolio managers, investment professionals and engineers on an AI-enabled research and decision platform. You will own high-impact projects from problem definition through model development, deployment and monitoring.

The role requires deep scientific depth and strong engineering judgement, with a practical understanding of how data and AI can improve investment research and decision-making.

Qualifications

• PhD in Computer Science, Machine Learning, Artificial Intelligence, Statistics, Applied Mathematics, Engineering, Physics or another highly quantitative field.
• Minimum 2 years of professional, full-time experience in data science, machine learning, applied AI or a closely related role. Doctoral research alone does not replace the professional-experience requirement.
• Strong Python proficiency and experience with core scientific and machine-learning libraries such as Pandas, NumPy, scikit-learn and PyTorch or equivalent frameworks.
• Strong grounding in machine-learning and statistical fundamentals, including problem framing, experimental design, validation, metrics, feature engineering, overfitting and uncertainty.
• Demonstrated experience delivering at least one end-to-end model or data product used by real stakeholders, from initial scoping through deployment and monitoring.
• Practical experience with LLMs and modern NLP, including retrieval-augmented generation, embeddings, vector search, prompt or context design and systematic evaluation.
• Proficiency in SQL and experience working with relational, columnar or document-oriented data systems.
• Ability to work with messy, incomplete and heterogeneous data while maintaining strong standards for data quality, testing, reproducibility and documentation.
• Strong written and verbal communication skills, with professional fluency in English and Mandarin.

Responsibilities

• Translate investment and research questions into well-defined data-science problems, measurable objectives and practical technical solutions.
• Develop and deploy machine-learning, statistical and LLM-enabled models for company research, industry analysis, market monitoring, event detection and knowledge discovery.
• Build robust workflows across the full data lifecycle, including data sourcing, cleaning, transformation, feature engineering, quality checks, modelling and monitoring.
• Develop retrieval, search and knowledge systems using structured and unstructured data, with rigorous source attribution and evaluation.
• Design experiments and evaluation frameworks covering model quality, factual accuracy, robustness, latency, cost and user impact.
• Work with engineers to productionize models and analytical tools through APIs, batch pipelines and monitored applications.
• Partner closely with investment users to understand workflows, communicate trade-offs and iterate based on evidence and feedback.
• Identify promising models, research and open-source technologies, and determine when they are—or are not—appropriate for real investment use cases.
• Improve tooling, documentation and processes to increase reliability, reduce manual work and enable reuse across the team.

Skills

Python Proficiency

Experiment Design

End-to-end ML

NLP & LLMs

Data Quality

English Mandarin

Education

PhD in Quantitative Field

Tools

Pandas

NumPy

scikit-learn

PyTorch

Job description

About The Team And Role

The investment team is building an AI-enabled research and decision platform that brings together proprietary knowledge, public information, market and alternative data, analytical tools and modern machine-learning capabilities.

We are seeking an AI Data Scientist to work directly with portfolio managers, investment professionals and engineers. The role will own high-impact projects from problem definition through model development, deployment, evaluation and ongoing improvement. The successful candidate will combine scientific depth with strong engineering judgement and a practical understanding of how data and AI can improve investment research and decision-making.

Principal Responsibilities

• Translate investment and research questions into well-defined data-science problems, measurable objectives and practical technical solutions.
• Develop and deploy machine-learning, statistical and LLM-enabled models for company research, industry analysis, market monitoring, event detection and knowledge discovery.
• Build robust workflows across the full data lifecycle, including data sourcing, cleaning, transformation, feature engineering, quality checks, modelling and monitoring.
• Develop retrieval, search and knowledge systems using structured and unstructured data, with rigorous source attribution and evaluation.
• Design experiments and evaluation frameworks covering model quality, factual accuracy, robustness, latency, cost and user impact.
• Work with engineers to productionize models and analytical tools through APIs, batch pipelines and monitored applications.
• Partner closely with investment users to understand workflows, communicate trade-offs and iterate based on evidence and feedback.
• Identify promising models, research and open-source technologies, and determine when they are—or are not—appropriate for real investment use cases.
• Improve tooling, documentation and processes to increase reliability, reduce manual work and enable reuse across the team.

Qualifications / Skills Required

• PhD in Computer Science, Machine Learning, Artificial Intelligence, Statistics, Applied Mathematics, Engineering, Physics or another highly quantitative field.
• Minimum 2 years of professional, full-time experience in data science, machine learning, applied AI or a closely related role. Doctoral research alone does not replace the professional-experience requirement.
• Strong Python proficiency and experience with core scientific and machine-learning libraries such as Pandas, NumPy, scikit-learn and PyTorch or equivalent frameworks.
• Strong grounding in machine-learning and statistical fundamentals, including problem framing, experimental design, validation, metrics, feature engineering, overfitting and uncertainty.
• Demonstrated experience delivering at least one end-to-end model or data product used by real stakeholders, from initial scoping through deployment and monitoring.
• Practical experience with LLMs and modern NLP, including retrieval-augmented generation, embeddings, vector search, prompt or context design and systematic evaluation.
• Proficiency in SQL and experience working with relational, columnar or document-oriented data systems.
• Ability to work with messy, incomplete and heterogeneous data while maintaining strong standards for data quality, testing, reproducibility and documentation.
• Strong written and verbal communication skills, with professional fluency in English and Mandarin.

PREFERRED QUALIFICATIONS

• Experience with financial, market, regulatory or alternative datasets, or with research-intensive decision environments.
• Experience building data or AI products in cloud environments and deploying APIs, batch jobs, monitoring or feedback loops.
• Familiarity with knowledge graphs, document processing, browser automation, data visualization or time-series and event-driven modelling.
• Evidence of technical depth through publications, patents, open-source contributions or substantial production projects.
• Genuine interest in companies, industries and investing; prior investment experience is valued but not required.

HOW WE WORK

• Ownership: Scope work clearly, set realistic milestones, communicate risks early and follow through on outcomes.
• Scientific rigour: Prefer measurable evidence, reproducible analysis and honest uncertainty over impressive demonstrations.
• Practical judgement: Start with the simplest viable approach, use advanced methods where they add value and understand when not to use AI.
• Collaboration: Work closely with investment and technology colleagues, seek feedback and communicate complex ideas clearly.
• Continuous improvement: Track developments in models, research and open-source tooling, and translate relevant advances into durable capabilities.