Staff ML Engineer: End-to-End Recommender Architect

Wonder-c9247c0; New York (NY); Hybride; USD 209 000 - 217 000; Plein temps

Avantages offerts par ce poste

Equity

401(k)

Medical/Dental/Vision

Paid time off

Parental leave

Discounted meals

Résumé du poste

Wonder, based in New York, seeks a hands-on senior staff engineer to own the architecture of its ranking and recommendation stack, spanning candidate retrieval, multi-objective ranking, and calibrated systems. You will drive cross-system design, guide evaluation pipelines, and lead offline/online experiments with cross-functional partners.

You will mentor engineers, translate trade-offs for executives, and push forward state-of-the-art IR and recommender research into production.

Qualifications

• 6+ years building and shipping ML systems at scale.
• 3+ years operating with senior-level scope across teams or systems.
• Deep experience in recommendation or information retrieval at scale.
• Production DL in TensorFlow or PyTorch including training, serving, and tuning.

Responsabilités

• Own the architecture of ranking and recommendation stack end to end.
• Lead cross-team design decisions and advocate for best engineering practices.
• Mentor engineers through design reviews and feedback sessions.
• Collaborate with product, data platform, and engineering partners to surface data and infra needs.

Connaissances

ML system design

Technical leadership

Information retrieval

Recommendation systems

TensorFlow/PyTorch

Cloud ML infra

Experimentation

Formation

MS/PhD in quantitative field

Outils

PySpark

Hive/SQL

AWS SageMaker

Description du poste

Wonder, based in New York, seeks a hands-on senior staff engineer to own the architecture of its ranking and recommendation stack, spanning candidate retrieval, multi-objective ranking, and calibrated systems. You will drive cross-system design, guide evaluation pipelines, and lead offline/online experiments with cross-functional partners.

You will mentor engineers, translate trade-offs for executives, and push forward state-of-the-art IR and recommender research into production.