AI Research Scientist(6239)

Position Overview

Taiwan Semiconductor Manufacturing Company (TSMC) is seeking applications for skilled Artificial Intelligence (AI) Research Scientists for their Artificial Intelligence for Business Intelligence (AI4BI) Center. Applicants should have graduated with a Ph.D. in the past 1-2 years in computer science, information systems, information science, statistics, systems and industrial engineering, or a related AI or machine learning field. This role, based in the San Jose office in a hybrid environment (four days in-office), is integral to TSMC’s global team. The selected candidate will design and implement advanced machine learning, deep learning, text mining, Large Language Model (LLM), foundation model, and/or network science-based approaches to extract insights from structured and unstructured data sources for various business intelligence applications. Additionally, the role may involve developing time series modelling and reinforcement learning approaches related to fab tool productivity and wafer start strategies.

Key Responsibilities

• Create and manage advanced machine learning, deep learning, LLM, foundation models, time-series modelling, and reinforcement learning algorithms to develop data-driven solutions for complex business problems.
• Analyze and interpret complex data sets to provide actionable insights for enhanced decision-making by business stakeholders.
• Collaborate with cross-functional international teams to identify, prioritize, and deploy predictive models and other analytics solutions.
• Work with senior and junior data scientists, as well as data engineers, to establish efficient procedures for data gathering, retention, and analysis.
• Consistently assess and refine model performance to ensure accuracy and reliability while communicating findings via internal reports.
• Design and implement cutting-edge AI, machine learning, and data science tools and techniques based on feedback from senior data scientists.

Required Qualifications

• Ph.D. in computer science, information systems, information science, statistics, or a related AI or machine learning field in the past 1-2 years.
• 4-5 years of AI/ML experience.
• Proficiency in Python, GitHub, and Markdown for data wrangling, preprocessing, and extraction from both structured and unstructured data sources.
• Solid understanding and hands-on experience of classical machine learning algorithms and paradigms including supervised and unsupervised learning.
• Strong skills in deep learning implementation including data encoding, processing, and various learning paradigms.
• Ability to adapt machine learning or deep learning algorithms based on unique dataset characteristics and business needs.
• Demonstrated network science skills and experience.
• Experience with text analytics techniques such as named entity recognition, sentiment analysis, and topic modelling.
• Hands-on research experience with time series modeling and forecasting using statistical, machine learning, deep learning, and/or foundation model approaches.
• Experience in reinforcement learning (value-based, policy-based, actor-critic, or multi-objective) and in fine-tuning LLMs using methods like low rank adaptation and few-shot learning.
• Familiarity with prompt engineering on LLMs using reinforcement learning techniques such as Direct Preference Optimization (DPO) or Proximal Policy Optimization (PPO).
• Knowledge of deploying models with dashboarding tools (e.g., PowerBI), rapid prototyping frameworks (e.g., Streamlit, Dash), or JavaScript frameworks (e.g., React).
• Willingness to travel to Taiwan for 1-4 weeks each year for training, team building, and project coordination.

Preferred Qualifications

• Experience processing text in financial, accounting, and market analysis reports.
• Familiarity with multilingual text analysis packages like Stanza, Polyglot, or Textflint.
• Knowledge of graph embedding techniques using graph convolutional networks, graph attention networks, or graph transformers with libraries such as StellarGraph, PyG, or Deep Graph Library.
• Understanding of neural information retrieval methods including deep structured semantic models, entity resolution, and retrieval augmented generation (RAG).
• Experience with adversarial learning, knowledge distillation, and self-supervised learning paradigms.
• Exposure to working with foundation models for language (e.g., LLaMA) and time (e.g., TimesFM) as well as other data modalities.
• Publications in leading AI conferences (e.g., NeurIPS, ICLR, ICML, KDD) or journals (e.g., IEEE TKDE, ACM TOIS).

Benefits & Perks

• Compensation: Base salary typically between $150,000 and $230,000 per year.
• Total Compensation: Consists of market competitive pay, allowances, bonuses, and comprehensive benefits.
• Extensive development opportunities and programs.