Avathon
AI Scientist, Product Lifecycle Management (PLM)
$160K–$240K
Pleasanton, California, United States · Full-time · Posted Oct 19, 2026
- Experience
- An MS with 5+ years of relevant industry experience is an alternative to the stated PhD requirement; no years specified for PhD holders.
- Education
- PhD in Computer Science, Machine Learning, Mechanical Engineering, or a related field, or an MS with 5+ years of relevant industry experience.
About the job
Build and productionize AI models that help manufacturers analyze engineering data, assess engineering-change impacts, and search parts, drawings, and change history.
What you'll do
- Develop models for part similarity, classification and substitution, engineering-change impact, and data extraction from engineering documents.
- Apply graph learning and build retrieval and LLM tools for engineering search; use Normal Behavior Models to flag unusual change activity.
- Partner with engineering to deploy, evaluate, monitor, and explain models, and work with AI agents that support engineers and planners.
What they ask for
- PhD in Computer Science, Machine Learning, Mechanical Engineering, or a related field, or an MS with 5+ years of relevant industry experience.
- Strong machine-learning foundation and depth in at least one of graph learning, NLP / LLMs, or computer vision on documents.
- Experience applying ML to engineering or manufacturing data such as BOMs, CAD, drawings, or part catalogs.
- Strong Python and standard ML stack (PyTorch, Hugging Face, graph libraries); experience shipping models to production.
Nice to have
- Experience at a PLM or CAD vendor or an engineering-heavy manufacturer.
- Publications or patents in related areas.
Pay
$160k - $240k USD base salary
Tools
Python, PyTorch, Hugging Face, graph libraries, Computational Knowledge Graph
AI focus
Develop graph, language, vision, retrieval, and LLM models for engineering-data intelligence, change-impact analysis, document extraction, natural-language search, anomaly detection, and AI-assisted planning.
Model-written summary · the original posting is authoritative