ML Engineer, Ads — AI-Powered Ad Platform (Hybrid)

Socket.dev; San Francisco (CA); Hybrid; USD 165,000 - 230,000; Full time

Benefits offered by this job

Stock options

Hybrid work model

Competitive compensation

Job summary

Higgsfield AI is redefining AI-native advertising, building at the frontier of generative AI and ads systems. This hybrid role in the San Francisco Bay Area focuses on developing ML systems that power advertising products, including ranking, recommendation, targeting, and optimization, with a strong emphasis on generative AI and model evaluation.

You will implement prompt engineering, post-training, and production ML practices to scale creative generation and advertiser performance,

Qualifications

• Deep experience building machine learning systems for advertising (ads ranking, targeting, optimization).
• Strong understanding of ads systems including ranking, recommendation, targeting, bidding, and measurement.
• Hands-on with LLMs, multimodal models, or generative AI systems.
• Strong experience with prompt engineering and model evaluation.
• Experience with post-training techniques ( supervised fine-tuning, preference optimization, RL ).
• Strong software engineering fundamentals and shipping production ML systems.
• Ability to operate across research and engineering to turn experiments into production.

Responsibilities

• Build and improve ML systems powering advertising products (ranking, recommendation, targeting, prediction, optimization).
• Develop models to improve ad creative quality, relevance, personalization, and performance at scale.
• Connect generative models with real-world advertising signals to create feedback loops for improvement.
• Apply prompt engineering and post-training techniques for advertising and creative use cases.
• Work on fine-tuning and preference optimization for adapting foundation models to advertising objectives.
• Design and run experiments across creative generation, ranking, targeting, and delivery.

Skills

ML systems for ads

Ads systems knowledge

LLMs / multimodal models

Prompt engineering

Model evaluation

Post-training techniques

Production ML systems

Cross-functional collaboration

Job description

Higgsfield AI is redefining AI-native advertising, building at the frontier of generative AI and ads systems. This hybrid role in the San Francisco Bay Area focuses on developing ML systems that power advertising products, including ranking, recommendation, targeting, and optimization, with a strong emphasis on generative AI and model evaluation.

You will implement prompt engineering, post-training, and production ML practices to scale creative generation and advertiser performance,