Machine Learning Engineer 4

Adobe Adobe · Enterprise · Bangalore, India

Machine Learning Engineer at Adobe to design, build, and ship multi-modal creative intelligence systems at enterprise scale, analyzing images, video, email, and social ads to connect signals to campaign results, powering insights, recommendations, and agent tools. The role involves training models, building pipelines, and operating production ML systems.

What you'd actually do

  1. Design and train models for creative understanding across vision, video, and language
  2. Build labels and features from LLMs, subject-matter experts, and user activity data
  3. Deliver pipelines for ingestion, featurization, training, versioning, and inference at scale
  4. Validate models with offline benchmarks and online A/B tests before release
  5. Operate production ML on Spark, Kubernetes, and GPU/CPU environments

Skills

Required

  • Strong Python and PyTorch for model development and deployment
  • Deep learning experience across vision, video, NLP, or generative AI (LLM/VLM fine-tuning, embeddings, or RAG)
  • End-to-end ML systems: feature engineering, training pipelines, inference, and production monitoring
  • Distributed data processing and cloud platforms (Spark, Kubernetes, GCP, AWS, or Azure)
  • Clear communication with senior leaders on ML trade-offs and technical direction

Nice to have

  • Experience with recommendation systems or marketing and creative content platforms

What the JD emphasized

  • own how those systems are designed, built, and shipped
  • train and improve models yourself
  • multi-modal creative intelligence at enterprise scale
  • turning hard ML problems into production systems that real customers use
  • end-to-end ML systems: feature engineering, training pipelines, inference, and production monitoring
  • owning ambiguous, multi-team problems with little day-to-day direction

Other signals

  • builds models and pipelines that analyze images, video, email, and paid social ads
  • connect those signals to campaign results
  • power insights, recommendations, and agent tools
  • own how those systems are designed, built, and shipped
  • train and improve models yourself
  • shape the architecture for multi-modal creative intelligence at enterprise scale
  • turning hard ML problems into production systems that real customers use