ML Benchmarking & Partner Integration Engineer

Google LLC; Mountain View, New York (CA, NY); Hybrid; USD 174,000 - 232,000; Full time

Job summary

Google DeepMind is seeking a Forward Deployed Engineer for the Kaggle platform on a 12-month fixed-term contract. You will onboard external benchmarks to Kaggle, run joint evaluations, and synthesize signals to improve model performance.

You will collaborate with engineering, product, and partner teams across time zones, producing high-quality docs and sample implementations to advance Gemini's production capabilities.

Qualifications

• Bachelor's degree or equivalent practical experience.
• 3 years of experience testing, maintaining, or launching software products, and 1 year of experience with software design and architecture.
• Experience with Machine Learning systems or Large Language Models (LLMs).
• Experience in a customer-facing role, managing client relationships or external stakeholders.

Responsibilities

• Own the end-to-end onboarding of critical new benchmarks to the Kaggle platform.
• Drive joint evaluations and benchmarking, build and run systematic evals on partner-specific datasets to deliver high-fidelity performance signals directly to modeling teams.
• Author high-visibility case studies, developer blogs, and reference implementations to showcase Gemini's production capabilities.
• Build feedback loop tooling, collaborate with product engineering to develop tooling that aggregates and surfaces developer feedback patterns at scale.

Skills

Bachelor's degree

Software testing

Software design

Machine Learning systems

LLMs

Customer-facing

Education

Bachelor's degree or equivalent practical experience

Tools

Benchmarks

Developer tools

APIs/SDKs

Job description

Google DeepMind is seeking a Forward Deployed Engineer for the Kaggle platform on a 12-month fixed-term contract. You will onboard external benchmarks to Kaggle, run joint evaluations, and synthesize signals to improve model performance.

You will collaborate with engineering, product, and partner teams across time zones, producing high-quality docs and sample implementations to advance Gemini's production capabilities.