Citi Group Senior data scientist large language model/credit 

China, Hong Kong 

508812550

Responsibilities

• Drive AI/Machine Learning development and implementation especially around Generative AI Large Language Model for Wholesale Lending and Counterparty Credit Risk analytics.
• Work independently or collaboratively (depending on the circumstance) to help business stakeholders to identify analytical/technical opportunities to improve the existing process or address existing pain points.
• Work with large and complex data sets (both internal and external data) to evaluate, recommend, and support the implementation of credit risk strategies using a variety of programing and AI tools (e.g. SQL, Python, C, Spark, Sikilearn, Tensorflow, PyTorch, HuggingFace, Large Language Model etc.).
• Design and implement Early Warning Indicators for Credit Risk, Counterparty Risk, Portfolio Health, Covenant Monitoring using either traditional statistical approach or advanced AI/ML modeling, graph analytics approach.
• Design and implement statistical and risk-driven solution for Limits Management and Exception Management.
• Be actively involved in and responsible for elements of the design / implementation / coding of feature driven analytic models, including data structure / design to support the delivery of such.
• Responsible for all the Risk and Controls requirement in analytics or technical process such as documentation, testing, and ongoing performance monitoring.

Qualifications

• 6-10 years of working experience in AI/ML domain preferably in Finance/Risk analytics field
• Must have demonstrated ability in programing skills using Python/C/SQL/Spark for data processing and data analytics.
• Must have solid knowledge and implementation experience of basic AI/Machine Learning such as Logistics Regression, Support Vector Machine, Random Forrest, Isolation Forrest, XGBoost.
• Must have solid knowledge and implementation experience of Large Language Model pipeline: Pre-trained LLM (such as Llama, Mistral), GPT, RAG, Fine-tuning, Prompt Engineering. Vector DB, LangChain, LlamaIndex etc.
• Working knowledge of Graph Analytics and the relevant tools such as TigerGraph, Neo4j etc. will be a huge plus.
• Working experience in Wholesale Credit Risk (Lending or Counterparty) analytics is not a must but a plus.
• Excellent analytic ability and problem solving skills.
• Excellent communication and interpersonal skills, be organized, detail oriented, and adaptive to matrix work environment

• Bachelors/University degree, Master’s degree preferred

Time Type:

View the " " poster. View the .

View the .

View the