Senior Principal Machine Learning Engineer, Ad Platforms

Disney Disney · Media · Seattle, WA +2

Senior Principal Machine Learning Engineer for Ad Platforms at Disney. This role focuses on applying ML and AI patterns to guide ML and Research teams in creating scalable, performant models and pipelines. Responsibilities include reviewing designs, influencing roadmaps, exploring solutions, defining software/models in high throughput/low latency environments, and mentoring. Requires deep expertise in LLMs, fine-tuning, RAG, prompt engineering, embeddings, vector search, and tokenization, with experience in model optimization and inference.

What you'd actually do

  1. Exploring, researching, implementing proofs-of-concept, and proposing solutions that will reduce cost and overhead, improve maintainability, minimize the time features该take to be in production.
  2. Defining, reviewing, and documenting software and models in a high throughput, low latency environment.
  3. Mentoring and inspiring team members in all aspects of professional software development.
  4. Establishing shared technical standards and working practices across globally distributed ML teams.
  5. Reading requirements documentation from Product, translating into areas of work and partnering with team leads through execution as needed.

Skills

Required

  • Python
  • Java
  • SQL
  • TensorFlow
  • Pytorch
  • Hugging Face libraries
  • TensorRT
  • ONNX
  • DeepSpeed
  • LLMs
  • OpenAI GPT models
  • Claude
  • Gemini
  • Llama
  • Fine-tuning (SFT, PEFT, RLHF, adapters)
  • Prompt engineering
  • Retrieval-augmented generation (RAG)
  • Context management
  • Multi-agent systems
  • Embeddings
  • Vector search
  • Tokenization
  • Ad Tech Industry knowledge

Nice to have

  • PhD in Electrical Engineering, Computer Science, Mathematics, or a related technical field

What the JD emphasized

  • Deep expertise in LLMs and OpenAI GPT models, Claude, Gemini, Llama, and similar models.
  • Fine-tuning (SFT, PEFT, RLHF, adapters)
  • Prompt engineering, retrieval-augmented generation (RAG), context management, and multi-agent systems and MCPs
  • Embeddings, and vector search
  • Tokenization
  • Model optimization and inference (TensorRT, ONNX, DeepSpeed)

Other signals

  • MLOps
  • large-scale systems
  • production ML
  • LLMs