AI Scientist, Platform

Avathon•Pleasanton, CA

4d•Onsite

About The Position

Avathon is hiring an AI Scientist in Product Engineering to build and improve shared AI capabilities used across its industrial AI platform and applications in planning, logistics, product lifecycle management, asset performance, and machine vision. The platform's core capabilities include Normal Behavior Models, a Computational Knowledge Graph, a Model Hub for serving ML and LLM models, and AI agents. The role partners with platform engineering to bring models and research methods into production so improvements benefit multiple applications. The AI Scientist will develop normal-behavior and anomaly-detection methods for time-series and event data, improve data quality and fill gaps in industrial data, and evaluate and adapt foundation models for industrial use. The role also covers knowledge graph methods, AI agents with appropriate guardrails and human approval, and evaluation frameworks to determine when models and agents are ready to ship. The scientist will work with the Model Hub team on serving, monitoring, retraining, and cost, package methods as reusable platform services, and communicate methods and results to product teams, customers, and leadership. This is an onsite position based in the Bay Area, California. Candidates are expected to work from the Pleasanton office five days a week. The annual salary range is $160,000–$240,000, with pay depending on factors including location, job-related knowledge, skills, and experience.

Requirements

PhD in Computer Science, Machine Learning, Statistics, or a related field, or an MS with at least five years of relevant industry experience.
Strong foundation in machine learning, with depth in at least two of the following: time series and anomaly detection; knowledge graphs and graph learning; or LLMs and agents.
Experience building ML or AI capabilities used across many products or customers.
Strong Python skills and experience with the standard ML stack, including PyTorch, Hugging Face, and LLM frameworks.
Experience shipping models to production, not only working in notebooks.
Must be based in the Bay Area, California, and report to the Pleasanton office five days a week.

Nice To Haves

Experience at Palantir, C3 AI, Databricks, Google, AWS, Microsoft, or an industrial AI company.
Publications or patents in related areas.

Responsibilities

Improve Normal Behavior Models and anomaly detection for time-series and event data across industries.
Develop methods for data quality and gap filling for industrial sensor and business data.
Evaluate and adapt time-series and multimodal foundation models for industrial applications.
Develop methods for ontology discovery, entity resolution, and mapping raw data into the knowledge graph.
Build and evaluate AI agents that query the graph, call models, and recommend actions, with clear guardrails and human approval.
Design evaluation frameworks to determine when an agent or model is ready to ship.
Work with the Model Hub team on model serving, monitoring, retraining, and cost.
Package methods as reusable platform services for application teams.
Write up methods and results clearly for product teams, customers, and leadership.
Work with platform engineering to take models from research to production.