Software Engineer - ML, Creative Optimization Ads

Luma AI, Inc. • Palo Alto, CA

40d

About The Position

Luma's mission is to build multimodal AI to expand human imagination and capabilities. We believe that multimodality is critical for intelligence. To go beyond language models and build more aware, capable, and useful systems, the next step function change will come from vision. So, we are working on training and scaling up multimodal foundation models for systems that can see and understand, show and explain, and eventually interact with our world to effect change. We are building the founding engineering team for a new product focused on AI-powered ad creation. This is a zero-to-one initiative where you will operate with a high degree of autonomy in a fast-paced, ambiguous environment. As the founding ML Engineer, you will be the bridge to our research team, applying a deep, quantitative understanding of ad effectiveness to build the engine that optimizes AI-generated creative. You will be instrumental in defining the technical architecture and core infrastructure that powers our creative platform.

Requirements

• You have 5+ years of hands-on experience in a machine learning or optimization role within the ad space.
• You have hands-on experience with fine-tuning large vision-language models (VLMs) using frameworks like TensorFlow, PyTorch, or JAX.
• You possess strong programming skills in Python and have experience with distributed computing frameworks.
• You possess a strong intuition for creative advertising and a passion for understanding both the art and the science behind what makes an ad resonant.
• You have a deep understanding of the core metrics advertisers care about (e.g., brand awareness vs. clicks).
• You have hands-on experience with containerization and orchestration technologies like Docker and Kubernetes, and have deployed and managed models on cloud platforms such as AWS, GCP, or Azure.

Nice To Haves

• Familiarity with building or applying AI agents.
• Experience as a founding engineer or working on a 0-to-1 product.

Responsibilities

• Apply state-of-the-art techniques, including efficient finetuning methods (e.g., LoRA), to finetune vision-language models (VLMs), driving the generation of high-quality, engaging video and image ads.
• Translate the nuances of what makes an ad effective into quantitative signals that guide our AI models.
• Define and implement a comprehensive suite of evaluation metrics, including both offline metrics (e.g., FID, CLIP score) and online performance metrics (e.g., CTR, ROAS) to measure ad creative quality, relevance, and performance.
• Build scalable ML-driven APIs and backend systems for creative optimization, leveraging real-time performance data and feedback loops to continuously improve model output.
• Design, run, and analyze A/B experiments in a marketplace setting to validate optimization strategies.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Industry: Publishing Industries
• Education Level: No Education Listed
• Number of Employees: 51-100 employees