全职 Machine Learning Engineer 工作, 薪水, Fourtitude.asia Selangor 公司招聘中 - Ricebowl

We are looking for a Machine Learning Engineer to develop systems that classify records into meaningful categories, such as bank transactions, records and financial data/metrics. You will improve classification accuracy across diverse transaction, descriptions and languages, and deploy reliable solutions for production use.

Key Responsibilities

·      Build and maintain models that classify records, transactions into predefined categories and subcategories.

·      Process records based on descriptions (merchant names, merchant category codes, decriptions and other relevant attributes).

·      Develop text preprocessing, merchant normalization, feature engineering, and classification pipelines.

·      Collaborate with business teams to define category taxonomies and consistent labelling guidelines.

·      Prepare labelled datasets, resolve ambiguous labels, and improve data quality.

·      Address imbalanced categories, limited labelled data, unseen merchants, and changing transaction patterns.

·      Evaluate models using precision, recall, F1 scores, and confusion matrices, including results by category.

·      Establish confidence thresholds and fallback approaches for uncertain predictions, including human review.

·      Deploy and monitor classification services, balancing accuracy, latency, and operating cost.

·      Incorporate user corrections and reviewer feedback into model improvement workflows.

·      Apply privacy and security controls when handling sensitive financial data.

Required Qualifications

·      Diploma/Bachelor’s degree in Computer Science, Engineering, Mathematics, Statistics, or a related field, or equivalent practical experience.

·      1 years of relevant experience in machine learning or applied data science.

·      Strong Python and SQL skills.

·      Hands-on experience with supervised classification and natural language processing.

·      Proficiency with tools such as scikit-learn, PyTorch, TensorFlow, or gradient boosting libraries.

·      Experience working with noisy text, imbalanced datasets, and multiclass or hierarchical classification.

·      Ability to design evaluation datasets that prevent data leakage and assess performance on unseen merchants and future transactions.

·      Experience deploying models and building reproducible training and inference pipelines.

·      Strong communication skills and the ability to explain classification errors and tradeoffs to business stakeholders.

·      Cleanse and standardize raw records and transaction data, including handling missing values, duplicates, inconsistent formats, incorrect data types, and noisy descriptions.

·      Develop automated data quality checks for completeness, accuracy, validity, consistency, and uniqueness across training and production datasets.

·      Investigate data quality issues, identify their root causes, and collaborate with data owners to implement corrections.

·      Monitor input data quality and detect changes that may affect classification accuracy.

Preferred Qualifications

·      Experience with banking, payments, personal finance, or transaction enrichment data.

·      Familiarity with merchant identification, entity resolution, and multilingual transaction descriptions.

·      Experience combining business rules, merchant mappings, and machine learning models.

·      Experience with embeddings, transformer models, active learning, or weak supervision.

·      Familiarity with MLOps, model monitoring, and cloud deployment.

·      Experience in data cleansing, data validation, and data quality checking using Python and SQL.