Data Scientist II

You will be joining a central team of talented Data Scientists, working on a wide variety of impactful & important problem areas like: Search, Personalisation, Content Moderation, Trust & Safety, AI led Automation, Computer Vision, Conversation AI (NLP) etc.
What you'll do
Ship and improve models in production. End-to-end ML model development lifecycle: data wrangling, feature engineering, model selection / architecture, training, offline & online evaluations, planning A/B experimentation & productionization.
Turn ad-hoc business needs into solutions, whether that means adapting an existing service or building a new feature.
Own your data pipelines: Data preparation, feature processing, training and deployment pipelines your models need.
Run and read A/B tests to decide what ships, and explain honestly when a result is noise.
Use AI tools as a core part of how you work, to move faster and to raise the quality of what you ship.
Work across any of the following domains: Search, Recommendations, Sell Form, Listing Understanding, Trust & Safety
Partner closely with Product Managers & other engineers in identifying potential data science solutions to important product / business problems.
Work closely with other data scientists, contributing to the culture of continuous learning & sharing.
Build multimodal understanding for the sell form, such as extracting attributes from listing photos and text and auto-filling listing details.
Improve search and recommendations, from candidate generation and semantic retrieval to ranking and relevance evaluation.
Work hands-on with Chinese-language data. Analyse listing text and search queries from Hong Kong and Taiwan directly, in Traditional Chinese and Cantonese as well as English.
Requirements
3+ years of applied ML engineering or data science experience, with hands-on experience of solving multiple product & business problems leveraging machine learning.
Working proficiency in English and in Mandarin or Cantonese. This role involves hands-on analysis of Chinese-language listing text and search queries from Hong Kong and Taiwan, so you'll need to read and reason about that data yourself.
Minimum Bachelor's degree in Computer Science, Statistics, Applied Mathematics or a related quantitative discipline, or equivalent demonstrated experience.
Solid foundation in maths, programming and ML/DL fundamentals.
Good programming ability in Python, SQL and experience with machine learning frameworks and libraries (e.g. PyTorch, Tensorflow, Keras, Spark MLlib, Sklearn etc.)
Good to have:
Experience in e-commerce, classifieds or C2C marketplaces, especially with cold-start listings, supply quality or long-tail queries.
Experience in one or more of: search, recommendations, marketplace trust & safety, or pricing.
Multimodal embeddings and vector retrieval, such as CLIP-style architectures and approximate nearest-neighbour search.
Attribute extraction or auto-fill from images and unstructured text.
A Master's degree in Computer Science or a related field.
Familiarity with Carousell as a user. Knowing the product first-hand helps a lot when tuning feeds and search quality.

Employment type
Full-time

Work arrangement
Hybrid