Machine Learning Engineer, Marketplace

Mercor•San Francisco, CA

20d•$130,000 - $500,000•Onsite

About The Position

As a Machine Learning Engineer on the Marketplace team, you will build the models and decision systems that power Mercor’s hiring engine. This includes search and ranking, candidate-job matching, marketplace recommendations, personalization, and allocation decisions across a rapidly growing talent network. This is an applied ML role with direct product and revenue impact. You will work on problems shaped by real marketplace constraints: sparse and delayed labels, cold start, noisy feedback, heterogeneous supply and demand, and the need to optimize across speed, quality, and conversion simultaneously.

Requirements

Strong track record of shipping ML systems into production
Experience with ranking, recommendation, search, matching, or marketplace problems
Good judgment on model design, objective functions, evaluation, and tradeoffs
Comfort working across the full applied ML stack: data, features, training, inference, and iteration
Strong engineering fundamentals and a bias toward simple, robust systems
Python, Go, embeddings, fine-tuning, RAG, Kafka, Postgres, Redis, Elasticsearch, Kubernetes, Terraform

Responsibilities

Build the models and decision systems that power Mercor’s hiring engine, including search and ranking, candidate-job matching, marketplace recommendations, personalization, and allocation decisions.
Develop ranking and matching systems that determine which candidates and opportunities are surfaced.
Create models for recommendation, personalization, and marketplace optimization.
Build retrieval, scoring, and decision pipelines operating at global scale.
Implement feedback loops that learn from downstream hiring outcomes, not just top-of-funnel engagement.
Develop real-time and batch inference systems embedded in product-critical workflows.
Improve candidate-job matching using embeddings, structured attributes, and behavioral signals.
Optimize ranking toward long-term hiring outcomes under delayed and incomplete labels.
Design models that balance marketplace objectives such as fill rate, quality, speed, and conversion.
Build systems for candidate allocation, opportunity routing, and liquidity optimization.
Develop evaluation and experimentation frameworks that connect model performance to business results.