Machine Learning Engineer - ML Systems

Clear Demand; United States; Remote; USD 120,000 - 190,000; Full time

Job summary

Clear Demand is seeking a technically skilled ML engineer to advance model-driven systems for product matching, classification, and similarity search. You will design and maintain production-ready inference pipelines, evaluate model performance, and collaborate with data scientists to translate research into scalable solutions.

The role emphasizes experimentation, benchmarking, and improving explainability, with opportunities to impact real-time decisioning and downstream product workflows.

Qualifications

• Proficient in ML fundamentals including classification, embeddings, similarity search, and evaluation metrics.
• Experience with ML libraries such as scikit-learn, PyTorch, TensorFlow, Hugging Face, or similar frameworks.
• Strong analytical skills with ability to derive insights from data and identify improvement opportunities.
• Experience with production ML workflows and inference pipelines.
• Good communication to explain model behavior and trade-offs.

Responsibilities

• Build, improve, and maintain ML inference pipelines and model-serving workflows.
• Work on model evaluation, error analysis, performance tracking, and experimentation.
• Analyze model outputs to identify patterns, failure modes, precision/recall trade-offs, and improvement opportunities.
• Collaborate with Data Scientists to operationalize models into production-ready systems.
• Support model training data preparation, labeling strategy, validation sample analysis, and benchmark creation.
• Work on ML systems involving embeddings, vector search, retrieval, classification, ranking, and match verdict generation.
• Develop analytical scripts, notebooks, dashboards, and reports to measure model performance and business impact.
• Help improve model explainability by surfacing signals, evidence, decision factors, and confidence indicators.
• Support automation workflows where model predictions are used for decisioning and downstream product workflows.
• Debug production ML issues related to model outputs, drift, thresholds, retrieval quality, and data inconsistencies.
• Contribute to building reusable ML utilities, evaluation frameworks, and experimentation workflows.

Skills

Programming

ML fundamentals

Data analysis

Model evaluation

Explainability

Communication

Python

SQL

Team collaboration

Tools

scikit-learn

PyTorch

TensorFlow

Hugging Face

sentence-transformers

Milvus

FAISS

Job description

Role Summary

This role is suited for a candidate with deeper interest and capability in machine learning systems, model development, model evaluation, inference workflows, experimentation, and analytical problem-solving. The candidate will work on model-driven systems that support product matching, classification, similarity search, explainability, automated decisioning, and quality improvement.

The role requires someone who can understand ML model behavior, analyze outputs, improve inference pipelines, measure performance, and collaborate with science and engineering teams to translate model capabilities into production workflows.

Key Responsibilities

• Build, improve, and maintain ML inference pipelines and model-serving workflows.
• Work on model evaluation, error analysis, performance tracking, and experimentation.
• Analyze model outputs to identify patterns, failure modes, precision / recall trade-offs, and improvement opportunities.
• Collaborate with Data Scientists to operationalize models into production-ready systems.
• Support model training data preparation, labeling strategy, validation sample analysis, and benchmark creation.
• Work on ML systems involving embeddings, vector search, retrieval, classification, ranking, and match verdict generation.
• Develop analytical scripts, notebooks, dashboards, and reports to measure model performance and business impact.
• Help improve model explainability by surfacing signals, evidence, decision factors, and confidence indicators.
• Support automation workflows where model predictions are used for decisioning and downstream product workflows.
• Debug production ML issues related to model outputs, drift, thresholds, retrieval quality, and data inconsistencies.
• Contribute to building reusable ML utilities, evaluation frameworks, and experimentation workflows.

Required Skills

• Strong programming skills.
• Good understanding of machine learning fundamentals, including classification, embeddings, similarity search, evaluation metrics, and model validation.
• Experience with ML libraries such as scikit-learn, PyTorch, TensorFlow, Hugging Face, sentence-transformers, or similar frameworks.
• Ability to perform model output analysis, error analysis, and metric-driven evaluation.
• Strong analytical skills with the ability to work with datasets, derive insights, and identify improvement opportunities.
• Good understanding of precision, recall, F1 score, confidence thresholds, false positives, false negatives, and sampling strategies.
• Strong in SQL , Python ML Libraries and data extraction for ML analysis.
• Familiarity with model inference pipelines and production ML workflows.
• Experience with LLMs, prompt-based evaluation, RAG systems, or agentic AI workflows.
• Ability to write clean, modular, maintainable code.
• Good communication skills to explain model behavior, analysis findings, and technical trade-offs.

Good to Have

• Experience with NLP, product matching, entity resolution, semantic search, or recommendation systems.
• Exposure to vector databases or retrieval systems such as Milvus, FAISS, OpenSearch, Pinecone, S3 Vectors, or similar technologies.
• Familiarity with ML Ops tools, model monitoring, experiment tracking, and deployment workflows.
• Experience working with e-commerce catalog data, product attributes, brand, size, pack, and taxonomy-related problems.
• Ability to build dashboards or analysis reports