Tasks:
- * Analyze model behavior edge cases and trade offs
- * Build and evaluate machine learning approaches for entity matching
- * Communicate experimental findings and recommendations to stakeholders
- * Compare LLM assisted approaches with lower cost alternatives
- * Create scoring and ranking methods for entity matching
- * Define benchmark datasets metrics baselines and error analysis processes
- * Design and run experiments to validate hypotheses
- * Develop embedding and LLM based matching approaches
- * Distinguish true matches from duplicates, lookalikes, and unrelated entities
- * Document successful and unsuccessful experiments
- * Establish experimental pipelines and research approaches
- * Optimize inference cost and scalability
- * Process messy, multilingual company and firmographic data
- * Retrain and maintain production classification models
Perks/Benefits:
Skills/Tech stack required:
[Baseline definition] [Benchmarking] [Candidate Generation] [Classification] [Clustering] [Deduplication] [Embeddings] [Entity Resolution] [Error Analysis] [Experimental Design] [Language Models] [Language Processing] [Large Language Models] [Machine Learning] [Machine Learning Inference] [Metrics definition] [Model Evaluation] [Natural Language] [Natural Language Processing] [Neural Networks] [PyCaret] [Python] [PyTorch] [Ranking] [Record Linkage] [Scalable machine learning] [Scalable machine learning inference] [Semantic similarity] [Set design] [Similarity scoring] [SQL] [Supervised Learning] [TensorFlow] [Test set] [Test Set Design] [Transformer Architectures] [Unsupervised Learning]
Educational requirements:
N/A
Role(s):
[Applied Data Scientist] [Data Scientist] [Engineer] [Learning Engineer] [Machine Learning Engineer] [Scientist]