Senior/Data Scientist

CHARLES & KEITH GROUP; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

CHARLES & KEITH GROUP is seeking a Senior Data Scientist to build and operate production ML on AWS, focusing on a contextual multi-armed bandit personalizing homepage, listings, and cart experiences with fast real-time endpoints. You will also drive generative AI applications for retail, including RAG over catalogs and agentic workflows.

The role spans demand forecasting, pricing, search, recommendations, and segmentation, with emphasis on production discipline, CI/CD for ML, and robust

Qualifications

• Bachelor’s degree in CS/ML/DS or related field.
• 4+ years of applied ML in production (7+ for Senior) with personalization/recSys at consumer scale.
• Hands-on AWS/SageMaker end-to-end (training jobs, pipelines, feature store, model monitor).
• Broader AWS stack: S3, Glue, Athena, Kinesis, Lambda, IAM, KMS; MLOps: CI/CD for ML, IaC, containers.

Responsibilities

• Contextual bandit personalization on AWS for flagship project (homepage, listings, product, cart).
• Engineer behavioral features from clickstream and warehouse data; design reward functions and exploration policies.
• Deliver end-to-end on SageMaker AI: training jobs, pipelines, feature store, model registry, real-time endpoints (<100 ms).
• Validate uplift via A/B experiments; transition ownership from vendor to production.

Skills

Python

SQL

PyTorch/TensorFlow

SageMaker

AWS

MLOps

AB testing

Reinforcement learning

Vector databases

LangChain

Hugging Face

TensorFlow

Education

Bachelor’s Degree in Computer Science / ML / DS

Tools

Amazon SageMaker

Hugging Face

LangChain

TensorFlow

PyTorch

Job description

We are seeking a Senior/Data Scientist with hands‑on experience on the AWS platform to build and operate production machine learning. The immediate priority is our real‑time hyper‑personalization engine - a contextual multi‑armed bandit built on Amazon SageMaker AI - and the role extends beyond it to generative AI and broader data science projects across the Group.

Responsibilities:

Contextual Bandit Personalization on AWS (Flagship Project)

• Design, build, and tune a contextual multi‑armed bandit personalizing homepage, listing, product, and cart pages to lift conversion rate and AOV
• Engineer behavioral features from clickstream and warehouse data, design reward functions, and tune exploration/exploitation policies per surface
• Deliver end‑to‑end on Amazon SageMaker AI — training jobs, Pipelines, Model Registry, Feature Store, Model Monitor, and real‑time endpoints (sub‑100 ms)
• Validate uplift through controlled A/B experimentation, and take the system over from the delivery vendor into production ownership after go‑live

Generative AI and Broader Data Science Projects

• Build production generative AI applications for retail - RAG over product catalogs and enterprise data, agentic workflows, content and service copilots - with evaluation, guardrails and cost control
• Deliver wider data science: demand forecasting, customer lifetime value, pricing and markdown, search, recommendations and segmentation

Engineering and Operations

• Build with production discipline: versioned pipelines, infrastructure‑as‑code, CI/CD for ML, containerization, security and cost control
• Monitoring, drift detection, retraining, and incident response

Requirements:

• Bachelor’s Degree in Computer Science, Machine Learning, Data Science, or related field
• 4+ years of applied ML in production for the Data Scientist level, or 7+ years for the Senior level, including personalization, recommendation or decisioning systems at consumer scale
• Hands‑on experience delivering machine learning on the AWS platform - Amazon SageMaker AI end‑to‑end (training jobs, Pipelines, Model Registry, Feature Store, Model Monitor, real‑time endpoints)
• Broader AWS stack (S3, Glue, Athena, Kinesis, Lambda, Step Functions, IAM, KMS) plus MLOps: CI/CD for ML, IaC, containers, and observability
• Contextual bandits or reinforcement learning (LinUCB, Thompson Sampling): reward design, exploration, cold start, and off‑policy evaluation; strong recommender‑system depth also considered
• Practical generative AI experience (prompting, RAG, fine‑tuning, evaluation, guardrails); Python and SQL, PyTorch/TensorFlow, Hugging Face, LangChain, and vector databases
• Rigorous A/B testing practice and excellent communication across business and technical teams
• Fashion retail or retail/ECommerce background, fluent in retail metrics and processes (conversion funnel, AOV, merchandising, seasonality) is an advantage
• Good to have: AWS Certified Machine Learning - Specialty or ML Engineer – Associate; Amplitude and Salesforce Commerce Cloud familiarity