Production ML Engineer – Ranking & Search Systems

Mendable; San Francisco (CA); Hybrid; USD 210,000 - 240,000; Full time

Benefits offered by this job

Salary that makes sense — $210,000–$0?

Competitive equity — details shared

Generous PTO — 15 days; 24+ days with

Parental leave — 12 weeks fully paid

Wellness stipend — $100/month

Learning & Development — up to $1,000/

Team offsites

Sabbatical — 3 paid months after 4 yrs

Job summary

Firecrawl is seeking an ML engineer to build the ML behind Firecrawl—the models and systems that serve them. You’ll train and ship ranking and relevance models for our search product and extend this work to extraction quality and LLM‑driven features.

You’ll own how we measure success via A/B testing and experimentation frameworks. The role requires 3+ years building ML or data-heavy systems in production; strong Python and production-quality code are expected, along with comfort in large-scale

Qualifications

• Shipped ML models into production and owned them after launch (deploy, monitor, retrain).
• Real ranking or relevance-modeling experience (learning-to-rank, recommendations, search quality).
• Comfortable in large, data-heavy systems: query logs, pipelines, large datasets.

Responsibilities

• Improve ranking and relevance for Firecrawl Search—from feature engineering to model training and production.
• Build and tune models for learning-to-rank, query understanding, and LLM-driven retrieval.
• Extend ML across Firecrawl's products: extraction quality, content classification, and evaluation of LLM-driven features.
• Mine query logs and behavioral data at scale to identify wins and failures.
• Build data pipelines turning crawl and query data into training data and features.
• Collaborate with platform, search engineers, and cloud DevOps to run models quickly and cheaply in production.
• Design and implement testing strategy: AB testing frameworks and offline evaluation.
• Partner on product launches: define success metrics, run experiments, decide ship/no-ship.

Skills

Production ML

Ranking modeling

Data-heavy systems

Python (production code)

A/B testing

Communication of results

Tools

MLflow

Kubernetes

Job description

Firecrawl is seeking an ML engineer to build the ML behind Firecrawl—the models and systems that serve them. You’ll train and ship ranking and relevance models for our search product and extend this work to extraction quality and LLM‑driven features.

You’ll own how we measure success via A/B testing and experimentation frameworks. The role requires 3+ years building ML or data-heavy systems in production; strong Python and production-quality code are expected, along with comfort in large-scale