Machine Learning Engineer
Adobe · San Jose, CA · Engineering and Product
About this role
Adobe is hiring a mid-level Machine Learning Engineer based in San Jose, CA. The posting calls out experience with Express, PyTorch, LLMs, System Design. Compensation is listed at $151,800–$265,350 per year.
- Role
- Machine Learning Engineer
- Function
- machine learning
- Level
- mid
- Track
- Individual contributor
- Employment
- Full-time
- Location
- San Jose, CA
- Department
- Engineering and Product
- Posted
- Mar 26, 2026
More roles at Adobe
Job description
from Adobe careersWe are building an advanced AI platform that powers next-generation creative workflows for flagship products like Photoshop and Lightroom. As a Sr. Machine Learning Engineer, you will combine hands-on engineering with architectural leadership to design and implement reasoning systems, tool orchestration, and multimodal integrations using cutting-edge large language models (LLMs) and vision-language models (VLLMs).
The Opportunity
• (M)LLM post-training and evaluation
• Agent system architecture design and implementation
• Practical agent development (Langchain, MCP, A2A)
• Image generation and editing
• Building scalable AI solutions
What you'll Do
• Developing next-generation AI agents that improve Adobe’s product ecosystem
• Crafting robust agent architectures that combine multiple AI capabilities
• Designing evaluation frameworks for agent performance
• Implementing state-of-the-art LLM techniques for specialized applications
What you need to succeed
• Master's/ PhD degree in Computer Science, Machine Learning, Data Science, or a related field
• Demonstrated experience in AI, LLMs, and agentic system development
• Proficiency in frameworks like PyTorch, LangChain, LangGraph, MCP, Agent Development Kit (ADK).
• Strong foundation in data structures, algorithms, and software engineering principles
• Excellent problem-solving and analytical skills, with a proactive approach to challenges
• Ability to move fast in an environment where things are sometimes loosely defined and may have competing priorities or deadlines