Machine Learning Engineer 5

Adobe Adobe · Enterprise · Noida, India

Machine Learning Engineer at Adobe Firefly’s Generative AI Services team, focusing on building and optimizing scalable generative AI systems for integration into Adobe products. Responsibilities include designing inference pipelines, optimizing models, building APIs, and collaborating on model training and serving.

What you'd actually do

  1. Design and evelopment of core GenAI services and APIs that integrate a wide range of generative models into Adobe’s flagship products.
  2. Design and build ML workflows for enterprise-scale model customization, serving, and ecosystem integration.
  3. Collaborate with Adobe Research and other model developer teams with a focus on model inference strategies and productization of those model
  4. Build and optimize GPU-accelerated pipelines for both (customized) model training and inference—prioritizing performance, scalability, and reliability.

Skills

Required

  • machine learning
  • production-scale deployments
  • large-scale, GPU-intensive GenAI systems
  • PyTorch
  • CUDA
  • Triton
  • TensorRT
  • Nvidia Dynamo
  • Python
  • generative model architectures
  • diffusion models
  • transformers
  • GANs

Nice to have

  • model serving
  • inference
  • orchestration
  • GPU resource management
  • Kubernetes
  • distributed systems
  • MLOps platforms

What the JD emphasized

  • production-scale deployments
  • large-scale, GPU-intensive GenAI systems (training, inference, and optimization)
  • PyTorch, CUDA, Triton, TensorRT, Nvidia Dynamo, and Python
  • diffusion models, transformers, and GANs
  • model serving, inference, orchestration, and GPU resource management in large-scale environments

Other signals

  • building scalable, high-performance generative AI systems
  • design and develop efficient inference pipelines
  • optimize models for latency and through at inference
  • build APIs and ecosystems that integrate generative models
  • productization of those models
  • GPU-accelerated pipelines for both (customized) model training and inference